“In what world,” New York Gov. Kathy Hochul asked on Friday, “would the police learn from a victim that she’d been literally gang raped while she was intoxicated by five men in a fraternity house and not elevate those allegations to the district attorney?” 

Hochul was speaking about the case of a former Cornell student who said she was raped by several members of the university’s Chi Phi fraternity in 2024. At the time, the district attorney filed no criminal charges.

The case has rocked the nation and forced a renewed reckoning over how sexual assault is dealt with in higher education. While the district attorney has defended his handling of the case, many politicians and legal experts are now examining whether certain details of the alleged victim’s account were neglected, and prosecutors have reopened the case.

Amy Wang is a staff reporter at the Washington Post who’s been covering the story over the past week. She spoke with Today, Explained co-host Noel King about what allegedly happened on the night in question, what happened afterward, and what we still don’t know. 

Since the two spoke, New York state Attorney General Letitia James has also been assigned to oversee the case and said it is now “an active criminal investigation.” As new information comes to light, further legal action could be taken. 

Below is an excerpt of the conversation between King and Wang, edited for length and clarity. There’s much more in the full podcast, so listen to Today, Explained wherever you get podcasts, including Apple Podcasts, Pandora, and Spotify.

In October of 2024, a young woman went to a fraternity house near Cornell University’s campus. Who is this young woman, and why did she go to the frat house that night?

The young woman is only identified in documents as Jane Doe, so we don’t know her name. She was, at the time, 20 years old. She was a sophomore at Cornell. And around 11 pm on the night of October 19, 2024, this Jane Doe goes to a bar on campus, and then, after that, goes to the Chi Phi fraternity house to hang out with a friend. 

She said that she was already pretty intoxicated at that point, and she’d had maybe 10 drinks in about three hours. Her friend at the fraternity house hands her a beer. After that is where she alleges that another member of the fraternity pressured her into taking drugs, specifically ketamine and marijuana, and drinking more alcohol. She alleges that two members of the fraternity assaulted her, and that one of them later sent a Snapchat message out to active Chi Phi members. The message is pretty graphic, but essentially invites them to join in, and several did.

In about three weeks, this young woman was going to go to the Cornell campus police and report what happened that night. But in the interim, what was going on? What was happening?

We now know, based on reporting by the New York Times and CBS News, this young woman was a member of a campus sorority, so she went to a member of her sorority and was trying to make sense of what happened. She was still processing it in real time. She wasn’t using, according to their reporting, the word “rape” or “assault.” And she was also seeing rumors circulating around campus about what had taken place at the fraternity house and was growing concerned about how they portrayed her. 

She, in the interim, also received a text message from one of the seven men at the fraternity essentially apologizing for how things went down the other night. She is not yet accusing them of assault, but is chatting, even seeks to reassure him in that first message. 

On October 22, she then makes an appointment at the campus health clinic because she had what she thought was a UTI and wanted to be tested because she told the doctor that she had gone through a traumatic experience. 

According to the New York Times reporting, you could see her thinking about what happened kind of evolve over those few weeks. She at some point says that this is starting to ruin her life, and she’s sleeping and crying and she’s feeling ashamed. She’s just really grappling with what had taken place, trying to make sense of it. By November 3 is when, according to these investigative documents that were obtained by the Times, she kind of comes to this conclusion that what happened at the party at Chi Phi was assault. 

She texts one of the seven fraternity members and says, “This was not consensual. I was given more ketamine than I can remember,” and that she was uncomfortable. A few days later, November 7th, 2024, according to these documents, she goes to the ER and asks for a rape kit. And the following day is when she goes to Cornell University Police.

When she goes to the police, how does she characterize what happened to the police? What does she tell them?

What she tells the police is that she believes she was assaulted, and she tells the police that it had started out as something that was consensual. And then at the end she says, “I can say with 100 percent confidence I was raped.” She tells police she was extremely uncomfortable and that she kept taking ketamine in part because she thought she might black out and pass out. She had thought of at least two of the men as her friends. After the events of that night, she is then confronted with rumors that were spreading across campus, sort of depicting her as loose and blaming her essentially.

What happens after Jane Doe talks to the police? What do they do?

When she talks to Cornell University Police, they summarize what she said in a six-page statement that she then signs. And that is the statement that is submitted to the Tompkins County district attorney. 

And the district attorney would be the person who would bring charges or decide not to bring charges. What does he decide? 

Yes, so the district attorney decides not to bring charges in 2024 based on, he has said, the sworn statement that she signs to police. 

The DA says no charges for the seven men. Were they disciplined in any way?

We don’t know exactly how each of the men were disciplined. Cornell has said that they faced a range of disciplinary measures, including expulsions and suspensions. One of the details that had gone particularly viral about this is that the disciplinary measures included for some of the students to write essays. Cornell has not denied that, but it has specifically denied that the students only had to write essays as a disciplinary measure. The university has cited privacy laws and said it really can’t release any more specifics about how individual students were punished or not punished. 

What did Jane Doe do after first talking to the police and then learning the DA wasn’t going to press charges?

After she goes to police, we know she remains on campus and she has to continue seeing some of these men. And she also participates in the school’s Title IX investigation, which takes place over several months and concludes at some point in 2025.

I want to ask you about the DA. He is out in public defending his decision to not press charges in 2024. How is he defending himself?

The DA is a big part of the story. This is the Tompkins County District Attorney. His name is Matthew Van Houten. And he has said that in 2024, based on the written statement from Jane Doe, they did not have what they needed to press criminal charges. 

It’s important to note that we, as in the public, we’ve not actually seen a copy of Jane Doe’s sworn statement. The DA issued a press release in which he summarizes that statement. And in that summary, he describes her characterization of this as voluntary and conscious and consensual. That is at odds with some of the investigative documents that we’ve seen reported since then, including that transcript of Jane Doe’s interview with Cornell Police in which she very confidently says that she believes she was raped and even says that she was physically hit during that night.

The DA said that he did not see, before the lawsuit, the screenshot of the Snapchat message where the frat members are talking pretty crudely about the situation, and he also says he didn’t see the transcript of the interview between Jane Doe and police.

We now know the names of the seven men who were there that night. Have they said anything?

The seven men themselves have not publicly said anything, and the lawyers for three of those seven men have publicly denied the allegations. One of them says that his client went into the room, but immediately left and didn’t participate in the assault.

This is a story that is developing very quickly, and it seems like new information is coming out just about every day. What do we know about what happens next?

So there’s at least two things going on here. There is the reopened criminal investigation. 

Separately, there’s the civil lawsuit that Jane Doe filed last month, and that is not dependent in any way on how the criminal investigation goes, but that will also proceed against the university, the seven fraternity members, and the fraternity itself.

There’s a lot to be anxious about as AI agents and bots get more integrated into every aspect of our lives. Meta, which recently agreed to pay over $17 billion to settle a lawsuit related to harming kids, now wants you to hand over all your online accounts to an inexplicably cute “agent” named Muse that might give a stranger your home address. OpenAI just introduced little blobs called Dots that can supposedly automate routine tasks, like planning dinner and ordering it from your favorite neighborhood spot. AI companies don’t just want you to talk to their chatbots; they want their agents to be an indispensable part of your daily routine.

As a parent, however, I think that perhaps the worst thing AI could do is be my kid’s friend. 

AI companions have been around longer than ChatGPT, and they’re growing in popularity with young users. More than half of teens have interacted with AI companions at least a few times a month, and one in three decided to talk about a serious issue with AI companions instead of real people, according to a Common Sense Media survey conducted last year. In certain groups and situations, a chatbot could offer some benefit — a lifeline for, say, LGBTQ youth who don’t have any in-person support networks, one study found. But even so, there are important caveats and risks.

Meanwhile, other experts are telling parents never to let their kids access AI companions, warning them that these chatbots have a history of offering dangerous advice and engaging in sexualized conversations. Sixteen-year-old Adam Raine died by suicide last April after developing a relationship and discussing his plans with ChatGPT.

If you have kids, these are horrifying stories to hear. It’s reasonable to wonder how widespread the issues are, and who they are affecting most. As of right now, the majority of younger kids aren’t using AI for companionship (phew!). But for those who are, where do the issues arise? And how do they go from using ChatGPT for help with their homework to opening up to it about their problems to preferring it to their human friends? 

Some answers are starting to emerge from research. 

For even young adolescents, there’s a direct link between AI companion use and mental health issues, according to a new study by researchers at the University of North Carolina at Chapel Hill. Their paper, which is still under review, is the first longitudinal study in the US that has looked at the bidirectional association between AI companionship in adolescents and mental health issues, according to one of the researchers. It charted AI use among 2,300 middle schoolers and found that about one in five used AI for emotional companionship. Those who used AI more reported higher levels of loneliness.

“It’s not just that the lonely kids are using AI, and they’re otherwise unharmed,” said Anne Maheux, an assistant professor of psychology and neuroscience at UNC Chapel Hill and the lead author on the study. “It is that using AI comes first, and middle school students are increasing their sense of loneliness following the use of AI.”

I recently spoke to Maheux about her research and the role of technology in young people’s lives. We discussed how much parents and teachers know about what is going on, what kids get out of human relationships that AI can’t provide, and whether she’d let a child under her care use an AI chatbot. While she said there is a lot to be scared of right now, there is some evidence that, when designed with the intention of promoting child learning and development and under the supervision of a caring adult, AI can offer some benefits, and, ultimately, she’s optimistic that we can design a better future for kids. 

The conversation below has been lightly edited for length and clarity.

How do so many kids end up with an AI companion? Do kids start using AI for homework help and end up in an emotional relationship with the technology? 

At a high level, AI companions — in our darkest vision of what that might look like — would be a kid calling an AI chatbot their best friend. That phenomenon is still quite rare. There are certainly plenty of kids who are experiencing that, but it’s not to say that it’s all kids or even most kids. I think that’s good news. 

But it’s also true that the line between opening up AI to ask a question about your homework and a full-blown AI companionship relationship involves a lot of gray areas. There’s evidence, mostly preliminary and anecdotal evidence, to suggest that very few people open up to AI with the primary initial goal of developing an intimate friendship or relationship. 

Most people open it out of a curiosity, sense of novelty, or seeking an answer to a question that’s very pragmatic. Then that can potentially evolve into something more akin to seeking personal advice or help making a decision or help making sense of one’s recent experiences with other humans. And that can potentially evolve to something that’s a bit more emotional.

What did your latest study on adolescents, AI companions, and mental health reveal?

Some kids seek out AI for certain purposes in different ways than other kids. For some kids, AI companionship is particularly attractive, especially the kids who maybe don’t have close friends or social networks in their offline life and who are already lonely. 

It’s a chicken-and-egg problem. Is it that these kids are seeking out AI companions or is it that having an AI companion is exacerbating those potential vulnerabilities or exacerbating risk for mental health problems? What we found is that interacting with an AI companion in a way that’s akin to emotional intimacy among middle school students is longitudinally associated with increases over time in loneliness. So the kids who are seeking out these companions, who are presumably already lonely, are also experiencing an increase in loneliness over time. 

There is this potential directionality question that we’re starting to answer here: It’s not just that the lonely kids are using AI, and they’re otherwise unharmed. It is that using AI comes first, and middle school students are increasing their sense of loneliness following the use of AI. In other words, middle school students using AI for emotional intimacy and companionship developed greater loneliness over time.

Do you have a sense of how often these kids are using the chatbots? 

In our sample — so again, these are US middle school kids — about 15 to 20 percent say that they’re using it at least a few times a week for any purpose. We don’t have specific rates of use for companionship because the questions that we asked were just not framed in a frequency measure, but more in how emotionally intimate they felt with the chatbot.

What else could have been causing that loneliness?

There are almost definitely many causes, and I think that our data suggests that AI companionship may be one cause. It certainly isn’t definitive. Obviously, I have to say that. I’m a scientist and trying to be hesitant about these things, but the reality is that whether or not AI use causes loneliness in this particular case, there are definitely other things that are also causing loneliness. 

That could be a myriad of things about the kid, maybe they have a lot of social anxiety, maybe they struggle with social skills, maybe the kid has had struggles in their family life and doesn’t have a strong sense of connection and belonging at home. It could be structural things. The kid just doesn’t have enough access to appropriate resources. So certainly there’s other things that are at play here, and typically technology use is one of many factors that can influence a kid’s mental health. 

And how much do you think that parents and teachers know about what’s going on, and how well-equipped do you think they are to intervene or to just be more involved in how these kids are using this technology? 

It’s a very slippery technology. The way kids are using it has changed rapidly in just the last year or two years. I would expect that parents, teachers, people who care about kids are probably struggling with the rapidity of change. But it’s also true that all the things that we have known for a long time about supporting kids and helping them thrive, those things are still true and they’ll be true with whatever technology comes after AI that kids are exploring. 

Those things include being curious, asking kids questions about their lived experiences online and offline, having conversations with kids so that they can understand the implications of different technology use. They can understand how, for example, AI is embedded within a broader system of social and political and economic factors, including the fact that many companies are trying to exploit kids’ attention in order to make money, which many kids are quite interested to hear about. 

Supporting kids also means providing a warm and supportive space for them to learn, to make mistakes, to be kids in the offline world, to have the frictionful, challenging, awkward conversations and first dates and all of the things that parents are sometimes trying to protect kids from by providing a safe container for them to explore those things. All of that can also provide avenues for kids to learn what they need to learn and develop — all the resources for a long life of thriving, even in the context of AI or other new technologies. 

What do kids learn from having these kinds of conversations with other humans that they can’t get from AI? 

Kids need to engage with other people, period. That’s true for all of us. We’re an incredibly social species. Simply being around other people and talking to strangers, let alone having strong social bonds, is really good for us. And all those things are uniquely important during adolescence, when kids have this neurobiological sensitivity to social information and reward. That’s when kids learn how to engage in conversation, they learn how to maintain eye contact, how to read body language, how to read emotions and perceive the world from the lens of someone else, to take their perspective. That’s not to mention other things, like being a member of one’s culture, what is valued, and what is expected. All of that is learned in the context of other human social relationships or relationships with broader social groups that a kid might be a member of. 

AI obviously can’t and never will be able to offer some of those things. It can’t actually experience empathy or care. It can’t engage in a true mutual human relationship. It may be able to simulate some of those things, and there’s some evidence that it can serve as a source of social practice. It can help kids work through the potential for difficult conversations and then serve as a bridge to engaging with other people. But there are aspects of AI that just simply can never replace the core sacred aspects of humans. 

Can you be more specific? What is it about human relationships that can’t AI simulate?

The way that AI systems are currently designed, which is not an inherent feature, but is a design choice, is to be sycophantic, highly agreeable, anthropomorphized, and all of those things can lead to kids developing expectations or understandings of interactive relationships that are actually fundamentally at odds with a normal human relationship. Some of the most consistent things that I’ve seen in the educated hypotheses on this topic and also in the emerging empirical research is that people need the friction that comes with human relationships. 

Human relationships are just hard and messy, and they’re also really beautiful. They’re what make so much of human life valuable and worth living. And with AI that’s always agreeable, that’s always available, that has no needs of its own, that can’t care or be cared for, it just doesn’t serve the same context to help kids learn things like conflict resolution or negotiation, being pro-social or caring for someone else who has needs. So many of those aspects that can feel uncomfortable in the moment, dealing with another person or a person who’s not responding right away, those things are actually important in the building blocks of developing the skills and capacities to be a functioning and thriving person in the social adult world. 

This is a problem with AI in general, isn’t it? It makes things effortless, but if you choose this path of least resistance, you’ll avoid the friction that causes you to learn and develop and build skills you need to exist in the world. 

Absolutely. And I think that all these things are connected too. AI offers this effortless, frictionless solution to our day-to-day problems, whether they’re social problems, whether they’re problems doing our homework, and those immediate solutions feel really good in the moment, so they’re quite tempting to keep coming back to. But there’s this real possibility that in the long term, that erodes our capacity for not only social relationships or cognitive skills, but just a general ability to withstand the momentary day-to-day distress of being a person living a normal life. And so I think every domain of human psychology comes together at that point, whether it’s social, emotional, mental health, or cognitive. 

Do you think that there are interventions that legislators could make or even that tech companies could make that would improve outcomes for kids, maybe make them a little less lonely? 

There are a million specific design changes that I would recommend. I think at a very minimum, AI systems can afford a pretty phenomenal amount of access to information, but there’s been no benefits that I’ve seen of them acting as friends. Kids need other human friends, full stop. Designing systems that are intended to exploit kids’ vulnerability or sensitivity to human social information and relationships, I think, is bad for kids. 

At the federal level, it seems that regulation is not coming, but at the state level, there have been many state regulations intended to protect kids, intended to create safer systems, to put in guardrails, to modulate different aspects of the models, the design, the user interface. These things could potentially make a real difference, I think.

What do you recommend parents do with regard to AI and their kids? Should they let them use chatbots? 

In short, no. But I am not a parent, and it’s much easier to dole out advice than it is to enact it. Parenting decisions about technology are not one-size-fits-all. Kids, parents, family dynamics, technology, and the context in which it is used differ, and in so many cases are personal choices (parents must weigh various potential opportunities and risks). Decades of research indicate what conditions a child needs to thrive. Extremely minimal research addresses the effects of AI on youth, and what we have so far suggests that the risks outweigh the benefits. 

I would not give a child under my care AI chatbots of any kind to use without me or another trusted adult present. I would talk to them openly about AI and its risks and benefits, but to me AI, like all technology, should be a tool used to promote human thriving, and when the evidence is unclear or suggestive of harms, I think it’s best not to experiment on kids. Currently, the off-the-shelf AI chatbots most kids are using appears more likely to undermine well-being for vulnerable users, especially children. 

Well, as a father of two daughters, I’m definitely terrified of what’s to come in the next few years. 

There’s a lot to be scared of, but there’s a lot to be hopeful of too. I think that everyone wants kids to thrive. Some of us are under different incentives, and some of those incentives might be perverse. But most of us are working really hard to design a better world for kids. 

This story was originally published in The Highlight. To get access to member-exclusive stories like this every month, become a Vox Member today.

In the moody alcoves of a lobby bar in Chicago in 2021 over French 75s, Dr. Louise King, a gynecologic surgeon, and Katie L. Watson, a bioethics professor, found themselves in conversation about a topic that had long frustrated them both: Female surgeons are consistently underpaid compared to male surgeons, even after accounting for specialty, rank, age, and other factors. 

What troubled them even more was how the pay gap compounded in King’s own field. Surgeons who perform procedures that are specifically for women, like gynecological surgeries, are paid less for their services than surgeons who perform comparable surgeries for the male anatomy. Gynecologic surgeons are also predominantly women. Over the course of the conversation, Watson termed this situation “double discrimination,” a phrase that captures the overlapping dilemma of lower pay for surgeries in a specialty that is overwhelmingly female and also serves a primarily female patient population.

That term, which has since entered the medical literature, stuck because the problem it describes has been well documented for decades. In 1997, a group of gynecologic surgeons published a paper in the journal Gynecologic Oncology with a title that gets straight to it — “Is Adam worth more than Eve?” Dr. Barbara A. Goff, now chair of Obstetrics and Gynecology at University of Washington Medicine, and colleagues found that doctors are paid around 44 percent more for performing male-specific procedures than analogous female-specific ones. 

In the three decades since, the problem hasn’t been fixed. To the contrary, it has been effectively institutionalized, built into the systems and codes that determine how surgeons bill for procedures of all kinds. Today, gynecologic or female-specific procedures are in large part still reimbursed at lower rates than male-specific procedures even when they take similar time and skill.

Those disparities have consequences that extend beyond the compensation of individual doctors. When gynecological procedures don’t bring in as much money to hospitals, those same hospitals are disincentivised to offer women’s health services. That can lead to worse care for women and, crucially, fewer training opportunities for the surgeons who do that work. 

But this historically persistent inequality is not a foregone conclusion.

How healthcare services are billed changes all the time as procedures get reevaluated, new ones get added, and medical societies and experts push for revisions. In fact, one of the biggest changes in how maternity care is billed and reimbursed is set to take effect early next year, splitting up services for prenatal care, delivery, postpartum care, and more. While this is a long-overdue fix for obstetrics, gynecologic surgery — which covers most of a woman’s life outside of childbearing years — continues to be left out of the conversation. 

And although real change will require concerted efforts, King and Watson — and other advocates who have spent years documenting this disparity — have a fairly clear sense of how to start fixing it.  

How surgical procedures are valued

The work of deciding how much a medical procedure is worth starts well above the head of any given doctor, medical practice, or hospital. In the United States, the Centers for Medicare and Medicaid Services, a federal agency that administers healthcare and health insurance policies, sets rates for different procedures based on a metric called relative value units (RVUs). An RVU is supposed to take into account how long a procedure takes, the skill it requires, and the intensity of work it demands. Both Medicare and private insurers consider these figures when they’re settling on reimbursement levels. If a procedure has a higher RVU, it is valued more highly, and makes more money.

In a recent study, cheekily titled “Price and Prejudice,” researchers from the Mount Sinai Health System found that out of 55 pairs of sex-specific procedures that they evaluated in 2023, 75 percent had lower RVUs for procedures on female patients. On average, RVUs for male-specific procedures were 30 percent higher, and when those procedures were performed in a hospital or surgical center, surgeons were paid around 26 percent more.

For example, doctors will often perform biopsies — taking a small sample of tissues, cells, or fluids — on both male and female genitalia to diagnose problems. In 2023, researchers found that Medicare’s standard national rate for reimbursing surgeons was $121.32 for a penile biopsy but only $65.74 for a vaginal biopsy. 

.ag-root{--ag-m: #3b3fb6;--ag-f: #e2552c;--ag-m-ink: #3b3fb6;--ag-f-ink: #c74d2a;--ag-bg: transparent;--ag-knock: #ffffff;--ag-card: #f5f4f1;--ag-ink: #16161d;--ag-ink-2: #4a4a55;--ag-ink-3: #686872;--ag-rule: #16161d;--ag-grid: #e4e3df;--ag-track: #ebeae6;--ag-focus: #1a73e8;--ag-head: var(--ag-host,Georgia,"Times New Roman",serif);--ag-ui: -apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue",Helvetica,Arial,sans-serif;--ag-head-weight: 700;--ag-head-track: -0.01em;--ag-ease: cubic-bezier(.2,.8,.2,1);container-type: inline-size;container-name: ag;display: block;position: relative;width: 100%;max-width: 760px;margin: 2.25em auto;padding: 14px 0 10px;border-top: 3px solid var(--ag-rule);background: var(--ag-bg);color: var(--ag-ink);font-family: var(--ag-ui);font-size: 16px;line-height: 1.4;text-align: left;-webkit-font-smoothing: antialiased;-webkit-text-size-adjust: 100%;box-sizing: border-box}.ag-root.ag-root{font-family: var(--ag-ui);font-size: 16px;line-height: 1.4;font-style: normal;font-weight: 400;letter-spacing: normal;word-spacing: normal;text-transform: none;text-indent: 0;text-align: left;color: var(--ag-ink)}.ag-root.ag-root i,.ag-root.ag-root em{font-style: normal}.ag-root.ag-root .ag-ft em{font-style: italic}.ag-root.ag-root *,.ag-root.ag-root *::before,.ag-root.ag-root *::after{box-sizing: border-box;letter-spacing: inherit;font-family: inherit;text-transform: inherit}.ag-root[data-ag-palette="teal"]{--ag-m: #00829b;--ag-f: #d4661a;--ag-m-ink: #007a91;--ag-f-ink: #b85a1a}.ag-root[data-ag-palette="violet"]{--ag-m: #5b3fb0;--ag-f: #c4661c;--ag-m-ink: #5b3fb0;--ag-f-ink: #b35e1c}.ag-root[data-ag-style="feature"]{--ag-head: "Bricolage Grotesque","Helvetica Neue",Arial,sans-serif;--ag-ui: "Instrument Sans","Helvetica Neue",Arial,sans-serif;--ag-head-weight: 800;--ag-head-track: -0.025em}.ag-root[data-ag-style="minimal"]{--ag-head: var(--ag-ui);--ag-head-weight: 650;--ag-head-track: -0.015em}.ag-root[data-ag-mode="dark"]{--ag-bg: #141417;--ag-knock: #141417;--ag-card: #1f1f24;--ag-ink: #ededf2;--ag-ink-2: #b9b9c4;--ag-ink-3: #8b8b97;--ag-rule: #ededf2;--ag-grid: #2c2c33;--ag-track: #26262c;--ag-focus: #8ab4f8;--ag-m: #6f73e8;--ag-f: #e5603a;--ag-m-ink: var(--ag-m);--ag-f-ink: var(--ag-f);color-scheme: dark}.ag-root[data-ag-mode="dark"][data-ag-palette="teal"]{--ag-m: #1e9db3;--ag-f: #dd7430;--ag-m-ink: #1e9db3;--ag-f-ink: #dd7430}.ag-root[data-ag-mode="dark"][data-ag-palette="violet"]{--ag-m: #8a6ee6;--ag-f: #cf7f22;--ag-m-ink: #8a6ee6;--ag-f-ink: #cf7f22}@media (prefers-color-scheme: dark){.ag-root[data-ag-mode="auto"]{--ag-bg: #141417;--ag-knock: #141417;--ag-card: #1f1f24;--ag-ink: #ededf2;--ag-ink-2: #b9b9c4;--ag-ink-3: #8b8b97;--ag-rule: #ededf2;--ag-grid: #2c2c33;--ag-track: #26262c;--ag-focus: #8ab4f8;--ag-m: #6f73e8;--ag-f: #e5603a;--ag-m-ink: var(--ag-m);--ag-f-ink: var(--ag-f);color-scheme: dark}.ag-root[data-ag-mode="auto"][data-ag-palette="teal"]{--ag-m: #1e9db3;--ag-f: #dd7430;--ag-m-ink: #1e9db3;--ag-f-ink: #dd7430}.ag-root[data-ag-mode="auto"][data-ag-palette="violet"]{--ag-m: #8a6ee6;--ag-f: #cf7f22;--ag-m-ink: #8a6ee6;--ag-f-ink: #cf7f22}}.ag-root[data-ag-mode="dark"]{padding-left: 16px;padding-right: 16px}@media (prefers-color-scheme: dark){.ag-root[data-ag-mode="auto"]{padding-left: 16px;padding-right: 16px}}.ag-root.ag-root .ag-x,.ag-root.ag-root p,.ag-root.ag-root h2,.ag-root.ag-root h3,.ag-root.ag-root button,.ag-root.ag-root select,.ag-root.ag-root input,.ag-root.ag-root ul,.ag-root.ag-root li,.ag-root.ag-root label,.ag-root.ag-root table{margin: 0;padding: 0;border: 0;background: none;box-shadow: none;font: inherit;color: inherit;letter-spacing: normal;text-transform: none;text-decoration: none;line-height: inherit;min-height: 0;width: auto;max-width: none;float: none}.ag-root.ag-root svg,.ag-root.ag-root .ag-pt-svgwrap svg{display: block;max-width: none;height: auto;overflow: visible}.ag-root.ag-root button{cursor: pointer;-webkit-appearance: none;appearance: none;border-radius: 0}.ag-root.ag-root :focus{outline: none}.ag-root.ag-root :focus-visible{outline: 2px solid var(--ag-focus);outline-offset: 2px;border-radius: 4px}.ag-root.ag-root .ag-sr{position: absolute !important;width: 1px !important;height: 1px !important;padding: 0 !important;margin: -1px !important;overflow: hidden !important;clip: rect(0 0 0 0) !important;white-space: nowrap !important;border: 0 !important}.ag-root.ag-root [hidden]{display: none !important}.ag-root.ag-root .ag-hed{font-family: var(--ag-head);font-size: 23px;line-height: 1.12;font-weight: var(--ag-head-weight);letter-spacing: var(--ag-head-track);color: var(--ag-ink);text-wrap: balance}.ag-root.ag-root .ag-dek{font-size: 15px;line-height: 1.4;color: var(--ag-ink-2);margin-top: 6px;text-wrap: pretty}@container ag (max-width: 480px){.ag-root.ag-root .ag-hed{font-size: 22px}.ag-root.ag-root .ag-dek{font-size: 14px}}.ag-root.ag-root .ag-ft{margin-top: 14px;padding-top: 12px;border-top: 1px solid var(--ag-grid);font-size: 12px;line-height: 1.45;color: var(--ag-ink-3)}.ag-root.ag-root .ag-method{margin-bottom: 8px;font-size: 12.5px;line-height: 1.5;color: var(--ag-ink-2);text-wrap: pretty}.ag-root.ag-root .ag-method b{font-weight: 600;color: var(--ag-ink)}.ag-root.ag-root .ag-ft-row{display: flex;flex-wrap: wrap;align-items: baseline;gap: 2px 14px}.ag-root.ag-root .ag-ft em{font-style: italic}.ag-root.ag-root .ag-credit{color: var(--ag-ink-3);white-space: nowrap}.ag-root.ag-root .ag-about-btn{margin-left: auto;padding: 4px 0;font-size: 12px;font-weight: 500;color: var(--ag-ink-2);display: inline-flex;align-items: center;gap: 6px;white-space: nowrap}.ag-root.ag-root .ag-about-btn::after{content: "";width: 5px;height: 5px;margin-top: -3px;border-right: 1.5px solid currentColor;border-bottom: 1.5px solid currentColor;transform: rotate(45deg)}.ag-root.ag-root .ag-about-btn:hover{color: var(--ag-ink)}.ag-root.ag-root .ag-ft-end{margin-left: auto;display: flex;align-items: center;gap: 14px}.ag-root.ag-root .ag-logo{flex: none;align-self: center;width: 31px;height: 15px;background-color: var(--ag-ink);-webkit-mask: url("data:image/png;base64,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") center / contain no-repeat;mask: url("data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAIQAAABACAQAAABxNBeOAAANcUlEQVR42tXbedRcZX0H8M+9d94kJITERBIoYASBtiyCpQcKLYgoHkpVVJCKylZ7WpYCttjWhUJdEE9BkKIeWpFTgZMeKBL2IOeAbMq+SAQEArIECCQEQhID78y9T/+Y5x1muffOwKE13Plv5j6/53m+v32ZxCW2kJhnhkJitKeQ4QXLrTVmgX+VKQSbWGAMmbnmIJcOrM018IJnZC51kkxunXjGZCZ7l+8KcmGETyG40ofMNVkm66KVashkxsx1oIdKKLYE9/q0ucY0rKPPFfGg9Z+W4OKR6M10bx8UTcFZxqzDz5jUh0cCopB7r8QkSaUqJRKTsKNmFxBNwdeRaIyshP/vT4KNrI6CX68Wy2wQV9Q/Ga7vgNsUnM26CkK3MVtuKcLQNbliRHATN0WKuYbbHC2Tj7DD7xCIINGybCQgpttgROrBkg71NQ6OamLdlogUz4+kQuuZMZJq6MBaSJ1gscaIsvQ7BSLBsyNIRCExa2QgNkYuc5+z1p2IYZiNaAMxCo/fObI0bB/3OF4+ktq9DYFIhipRbrJdMOYa16/b8tANRIhAjCLyc0akvJN5CsFJ1vmnF4jnB2TkzQOR4ACkrnHHui4P/ZC8y6tDQ6qm4AKGZAqJxFRPC4LdY3D1NnkSrG/pUCBagmuGXq2BTwkKN40oZesYGA8MzUFbgjuH2pIUNyoEf/H2koeJw/9saOKVCxbLaqHIJHZRKNwjXUflIelNGtM+IIZnGwlmWH+ok/17icSpiiFAJP1HeiuuNYThDWlka2km3MB3ojmszz9ftVmN5qcS77FW4eHKZD2RafSpTLusk7xBALIus52NoIQTb0wyw7TXmdtv+58dyahONtPTNYi3HG0KzjCuodW3Po1VjfYzxXSp1dYoYibSGDlDzeSRznSpV+TxfKHm7LlNHWRvW1pf02MucbZXe5nawEFDJaJtJfasNIGJxGzLFZ42rU8ekg7wG/mYky3wmCUWud8ST7rSyfY2tYdvwyKghn39h3st9YIHnW+fkni5e0XiK17qu8/dNhsUmg+MULlsCfavjCQaOF4Q/HPfO+3LzXCwK70caV1uV7PM9D6Xx28e902bREDr3XPiML8cON0Cc0qhSLGBq2Is1JIrFFpeE9xpvf5XtxkSRUyEVEdUAJFITLFYYZlZXZdJJdjQiZ6KVMYV/rNn7d2aURqXOa6Wswl2cUuUz2a8VK6lJXjIpgNrE6lpbon7Dt7nC/0aNNeqkWLLEyuAmFCv4Ftdb7Sd7TGejRLV0lJ40lSpRqfC+U+CpjyCcZkZpVC05eoETUGrRHrHBXeZ0mN22yp5Wfx1UNVz9wyawaeGKkdT8L0KIFKJ2xVe8XuSjh6zvZvj2qJD49geqBIf78QwRbzO7AEoGtjIQkFRGe+MC77TY2ca+FoFDO3PK4NW9Z6RgLiw1KBl2FMQfL8jBxkOt7oLhLbEvWR2l+pk2KcnmBsX/NzkHs42sLPHBWtrwr5CS8u2HUZk2F2ra//Bz5rBiywcGlu2BNdViu3lCmttKZHGo3y7s6obyv/ugTLDx/reGo9176wLhv290uO9qlk1v4sVUzxW4w0LoT8YaODcoQ40F9xboRbbGo/ZaRal4ZweWXgdyk91OdP2zp8d2Lkp+FAXraPj94td7Y4aW1YovGoeUhnOsMr8SuhaCpcOAvHNkYB4wqSBbCPD2YKmHToR3zklmlkIVpnTs76B4wZ2bincoxGvc5IguMIHTQZH1shuM7rvyRLbCA6SWqoohaIl+MQgEEeNFGSvGCjgphIbe1lwGRGGb5UaqJbg+j7VauCUkp3HBfuB0wXPOrAnmD6lEoqW4I5I9za30bFBeckON/fbuwZdtrsOiHGbl1zlREHhTzEpivp4Bbf+pc/rNHBBqWrcbmOpHwluMi9CkETdn2xxBZfbZ9waBwq2wxj+diCcGhes8Ptlxm7XoXFE+/c/6msGJNb3lML1sa23ldXyUkq5YI8+r9OuXwya1PvMwoWCi0wpAe+va2XiYCxzftwrw36e6Hvvwb6bdI6zpdZQKPKOEes/VPDnmCRxU8UR293T3hZRgjGP95izpuBes012teACyYDDTiSmeabCCDYFpzhSMLdTE8kw0zF+6jeWW+IGx5pWFggkmGXFSLHlp3v4k8jcp3CHtANKs5JTN5VUQjbx2659xwW32sB0Nwpu77jiQVU+rWKnQrDYuJMH3LSYP08tbWl0gMgsHimkOqYvhN5XEPwlMjM8J690VsGNfdfKsFsHhkJTsNCYOe4W5G6oyDsy7KSoYFvbu02NuUl3uJ12VUCS6oz91qHmsin4RhcQKa5VeCDOwXypxvPkgvtKOHt4XNMG8Ed4j0cErwkerJnESP2qgnEthf0rUvqBilZakqwuHakoMqeLL8FO9pL4rqZgmr8Tagt0G1pvIArZDrSkUl/yebu4xVZx4mpKTLvLZKJwJRXN5USrAsIwrC/fwPeHRhItwSUdrDOcp/AbU2Ux+2zV+pymrQc68dfKvSZ4yodxgDWRSi54siR8e1059qgxl2cM7cDUAPHVkYC4MV4glXi31YIvxvjh6prMMBfcIDikK8B+3WcE882I8cjE5XLBY5USkWCaZ0vNe0twy8h9+xptrXefi7pWfFuw1DukQysaLcFfWeK6PgvzXsE99sUmrhK6IpBccP+QYuzFpScuBCvNfnNQDKbDVUA8bUqUh3dYGsfEJuOTtasLwcbOEmwf441EQ2ZH+4PjvNi3viwg72fd0RWsG4x33lCT530jjZSttGEMXI8RvGTjWGc6fYjHeAZba7nHzL69D7VoIGFv0zq3RtNT7FjhQifm+BpvDojNhraCC0HLlkhN8pDgTKJ/vq5GIlqChUQHe79P2MRsW9nPDyzpuLzByxxfc5l2Xe3JUoNZFry9xa3gQrAzOECwxhaxYjDmsZpwrBlrmZNi5hg0re46dl4hRR+oFe8Ml5TCXwhW22gUK5GWzMOs8eLQxl9BNEPH4SKPSxWYUWuc2qXA9nDZl+1tgee1LPGIlkJWwrkg9bL7KiOFCaq/KD1xIjfN7pFJ/yet4KbgMIk9FcZtL4nX2Nx4rTTl/rDjdtt9rtkm19RAWgoLhwj3sFjix6OYy/JGyGiDp3MEX5C4wiJpbL1NqjFMBZ7zRKTdloBXveg1e9Y0kxOXDgGiwANWSUvOnGFv0+TDlKN8g2dHkp31zLWv4LSeQ4XaMbRHre0cOFfEnsYfV5wkaFjtMuSx/tgoUaAgscIvS9UnUdjY+4crRzkQz4wExDRHGHOdWzvywBqv1gLxQN+eiWBL8ypykxyXWyqTSGMvK1cMzFxkghdr5OWzwyd+G29SIlLsZS5O7RjHgBVWdPL8sudXA3QKu0i1KtpFnInMOLa2nRmWW+SJ2AmfgKHlUPsJpVzP8FFzvRDN+RuKLfcc+U8sd/ZUFlIqK1PljrCB/6owlS3B1TFo29nCTv3ztxbYrkMnldjCa+7s6ngMGsx/rOzV1vK63QouRmgGH9CzQQOnxsLKaCVfGh4uhb3QkttRgsMjCC3NCPIq+0ZKGW72vDmVBaVc4Unr9RVoDAu0Rm0F5woPGeshnuGDlccJlvTVIdqg55Xh8VngI52qVeiqYK3xBzGoP0ywDy6uYEFbtr7YJxNt8zl12ETMUyNN131+gHhissdLS+wtwe19wtiQOLJUMdrJ93SJWZaWpPXNqDYNM61xPhKHVCplIbfSuzunnZicOtijJSWiN9QKnhC3/nGO6kJdU/CTPguRYUHJ8dtt3N06WUmzopm/A86z2sxYAHilUo5zwW3W60xXMOZEwel1c1ejtIKbgn8oMUCJxMzS0m1T8O99lW9mWD5w+PZQwBFRYu6uKPM0BYfaQfC5TufrokrlaN/mZtuD2Q5yh+C8+smc4a3gQuEFM0uJZPhMSY+rKfhqn2lN7DcAePsqX4veYraXK7icC67yhJ91tW/qvV0uaLrD7ZYJgp9I66f4hreC6/P8LLrE8YE1R/WsyTC/b5+24fx6J6nfzNpKcW8POs7rOPBU4ue1ktwN0oVxVmdIkHVk7TxBYaWNSlsuE/Z4kuv7oGgKDu8p0CXmWNl1zSKOAx3VmWxgA8sqgRgXHNpldTK8f0iZsTCuGQdZho6oZkNawc1YiMlqPc80VwvyDpWm4G+6gGjgy51jT+j2r+3RV828sYtGPww/LJnbO7dmRKi9y7hju4zmECD+RMt4HPvq/TQ1rbV5pTy8foXUqZ3hsSI2hSb+ODuGd1oml2tFENY61QYDAdrnYpNn0JxePTCpm0itb1EpFBNDavfaNY6ejJRHbFkbQ5w3QobfFry9/KIjEa3e2TWXddn4V5xjm5JR01Tq2jhtV3SGCIPgYlNKhDvFPI9G9/r6irZyrXDCQE+9JtZOBNN9zzSh5PeAEzwiGSGFaSdGH3GYPc0G/+M0zxmzva/EUt9Kd7vcJZ6OfaswcL7pfuzjfUnhvzkznlRJGjfXD3yy7/vF5vuhJfGNN5J0vCXPRI64qT+zm21tq2GV1Ppetsiv3ekuT3ZS6aKCLXzGIbYz1SoPW2i+52tmrtsX/ajD7GSGtZ52l5+63uq6Oe//BeYA3+6agCy9AAAAAElFTkSuQmCC") center / contain no-repeat}.ag-root.ag-root .ag-about{margin-top: 8px;padding-left: 10px;border-left: 2px solid var(--ag-grid);display: grid;gap: 6px;font-size: 12.5px;line-height: 1.5;color: var(--ag-ink-2)}.ag-root.ag-root .ag-about a{color: inherit;text-decoration: underline;text-decoration-color: var(--ag-ink-3);text-underline-offset: 2px}.ag-root.ag-root .ag-about a:hover{text-decoration-color: currentColor}.ag-root.ag-root .ag-sw{display: inline-block;width: 10px;height: 10px;border-radius: 50%;flex: none;vertical-align: baseline}.ag-root.ag-root .ag-sw-m{background: var(--ag-m)}.ag-root.ag-root .ag-sw-f{background: var(--ag-f)}.ag-root.ag-root .ag-num{font-variant-numeric: tabular-nums}.ag-root.ag-root .ag-btn{display: inline-flex;align-items: center;justify-content: center;gap: 8px;min-height: 44px;padding: 10px 18px;border-radius: 999px;background: var(--ag-ink);color: var(--ag-knock);font-size: 15px;font-weight: 650;line-height: 1;transition: transform .15s var(--ag-ease),opacity .15s}.ag-root.ag-root .ag-btn:hover{opacity: .88}.ag-root.ag-root .ag-btn:active{transform: scale(.97)}.ag-root.ag-root .ag-btn[disabled]{opacity: .35;cursor: default}.ag-root.ag-root .ag-btn-ghost{background: transparent;color: var(--ag-ink);box-shadow: inset 0 0 0 1.5px var(--ag-grid)}.ag-root.ag-root .ag-btn-ghost:hover{box-shadow: inset 0 0 0 1.5px var(--ag-ink-3);opacity: 1}.ag-root.ag-root .ag-icon-btn{display: inline-flex;align-items: center;justify-content: center;width: 44px;height: 44px;flex: none;border-radius: 50%;color: var(--ag-ink);box-shadow: inset 0 0 0 1.5px var(--ag-grid);transition: box-shadow .15s}.ag-root.ag-root .ag-icon-btn:hover{box-shadow: inset 0 0 0 1.5px var(--ag-ink-3)}.ag-root.ag-root .ag-icon-btn svg{width: 18px;height: 18px}.ag-root.ag-root .ag-axis-cap{display: flex;justify-content: space-between;gap: 12px;font-size: 12px;color: var(--ag-ink-3)}.ag-root.ag-root .ag-bar-track{position: relative;height: 16px}.ag-root.ag-root .ag-bar{position: absolute;left: 0;top: 0;bottom: 0;width: 0;border-radius: 0 4px 4px 0;transition: width .9s var(--ag-ease)}.ag-root.ag-root .ag-bar-m{background: var(--ag-m)}.ag-root.ag-root .ag-bar-f{background: var(--ag-f)}.ag-root.ag-root .ag-bar-guess{background: repeating-linear-gradient(135deg,color-mix(in srgb,var(--ag-f) 55%,transparent) 0 2px,transparent 2px 7px);box-shadow: inset 0 0 0 1.5px var(--ag-f);transition: none}.ag-root.ag-root .ag-q-stage{margin-top: 12px}.ag-root.ag-root .ag-q-prog{display: flex;align-items: center;gap: 6px;margin-bottom: 8px;font-size: 12px;color: var(--ag-ink-3);font-weight: 600}.ag-root.ag-root .ag-q-prog span{margin-left: 6px}.ag-root.ag-root .ag-ink-m{color: var(--ag-m-ink)}.ag-root.ag-root .ag-ink-f{color: var(--ag-f-ink)}.ag-root.ag-root .ag-q-chart{margin-top: 10px;display: grid;gap: 8px}.ag-root.ag-root .ag-q-row{display: grid;grid-template-columns: 1fr;gap: 5px}.ag-root.ag-root .ag-q-row-hd{display: flex;justify-content: space-between;align-items: baseline;gap: 10px;font-size: 13px;color: var(--ag-ink-2)}.ag-root.ag-root .ag-q-row-hd b{font-weight: 650;color: var(--ag-ink);font-size: 15px;font-variant-numeric: tabular-nums}.ag-root.ag-root .ag-q-lab{display: inline-flex;align-items: center;gap: 6px;font-weight: 600}.ag-root.ag-root .ag-q-guess-tick{position: absolute;top: -5px;bottom: -5px;width: 2px;margin-left: -1px;background: var(--ag-ink);opacity: 0;transition: opacity .3s .6s}.ag-root.ag-root .ag-q-guess-tick.is-on{opacity: 1}.ag-root.ag-root .ag-q-guess-lab{position: absolute;top: 100%;margin-top: 7px;transform: translateX(-50%);white-space: nowrap;font-size: 12px;font-weight: 600;color: var(--ag-ink-2);opacity: 0;transition: opacity .3s .7s}.ag-root.ag-root .ag-q-guess-lab.is-on{opacity: 1}.ag-root.ag-root .ag-q-guess-lab.is-end{transform: translateX(-100%)}.ag-root.ag-root .ag-q-guess-lab.is-start{transform: none}.ag-root.ag-root .ag-range{display: block;width: 100%;height: 40px;margin-top: 2px;background: transparent;-webkit-appearance: none;appearance: none;cursor: grab;touch-action: pan-y}.ag-root.ag-root .ag-range:active{cursor: grabbing}.ag-root.ag-root .ag-range::-webkit-slider-runnable-track{height: 4px;border-radius: 2px;background: var(--ag-grid)}.ag-root.ag-root .ag-range::-moz-range-track{height: 4px;border-radius: 2px;background: var(--ag-grid)}.ag-root.ag-root .ag-range::-webkit-slider-thumb{-webkit-appearance: none;width: 28px;height: 28px;margin-top: -12px;border-radius: 50%;background: var(--ag-knock);border: 2px solid var(--ag-f);box-shadow: 0 1px 4px rgba(0,0,0,.18)}.ag-root.ag-root .ag-range::-moz-range-thumb{width: 24px;height: 24px;border-radius: 50%;background: var(--ag-knock);border: 2px solid var(--ag-f);box-shadow: 0 1px 4px rgba(0,0,0,.18)}.ag-root.ag-root .ag-range:focus-visible{outline: none}.ag-root.ag-root .ag-range:focus-visible::-webkit-slider-thumb{outline: 2px solid var(--ag-focus);outline-offset: 2px}.ag-root.ag-root .ag-q-hint{font-size: 12px;color: var(--ag-ink-3)}.ag-root.ag-root .ag-fade-in{animation: ag-in .5s var(--ag-ease) both}@keyframes ag-in{from{opacity: 0;transform: translateY(8px)}to{opacity: 1;transform: none}}.ag-root.ag-root .ag-q-end{text-align: left}.ag-root.ag-root .ag-q-recap{margin: 12px 0 4px;display: grid;gap: 6px}.ag-root.ag-root .ag-q-recap-row{display: grid;grid-template-columns: minmax(0,1fr) auto;gap: 2px 12px;font-size: 13px;color: var(--ag-ink-2);align-items: center}.ag-root.ag-root .ag-q-recap-row .ag-num{color: var(--ag-ink);font-weight: 600}.ag-root.ag-root .ag-q-mini{grid-column: 1 / -1;position: relative;height: 8px;border-radius: 4px;background: var(--ag-track)}.ag-root.ag-root .ag-q-mini i{position: absolute;top: -2px;width: 12px;height: 12px;margin-left: -6px;border-radius: 50%;border: 2px solid var(--ag-knock)}.ag-root.ag-root .ag-q-mini i.ag-g{background: var(--ag-knock);border: 2px solid var(--ag-ink-2);width: 12px;height: 12px}.ag-root.ag-root .ag-q-cta{display: flex;flex-wrap: wrap;gap: 10px;margin-top: 12px}.ag-root.ag-root a.ag-btn{text-decoration: none;color: var(--ag-knock)}.ag-root.ag-root .ag-pt-svgwrap{position: relative;margin-top: 8px;touch-action: manipulation}.ag-root.ag-root .ag-pt-svgwrap svg:focus-visible{outline: 2px solid var(--ag-focus);outline-offset: 4px}.ag-root.ag-root .ag-pt-dot{transition: transform .9s var(--ag-ease),opacity .5s,r .15s;cursor: pointer}.ag-root.ag-root .ag-pt-dot.is-m{fill: var(--ag-m)}.ag-root.ag-root .ag-pt-dot.is-f{fill: var(--ag-f)}.ag-root.ag-root .ag-pt-dot{stroke: var(--ag-knock);stroke-width: 2px}.ag-root.ag-root .ag-pt-dot.is-on{stroke: var(--ag-ink);stroke-width: 2.5px}.ag-root.ag-root .ag-pt-dim .ag-pt-dot:not(.is-on):not(.is-hit){opacity: .45}.ag-root.ag-root .ag-pt-grid{stroke: var(--ag-grid);stroke-width: 1px}.ag-root.ag-root .ag-pt-zero{stroke: var(--ag-ink);stroke-width: 1.5px}.ag-root.ag-root .ag-pt-tick{fill: var(--ag-ink-3);font-size: 11px;font-family: var(--ag-ui);font-variant-numeric: tabular-nums}.ag-root.ag-root .ag-pt-card{margin-top: 14px;min-height: 88px;padding-top: 14px;border-top: 1px solid var(--ag-grid);display: grid;gap: 8px}.ag-root.ag-root .ag-pt-empty{font-size: 14px;color: var(--ag-ink-3);align-self: center;justify-self: center;text-align: center}.ag-root.ag-root .ag-ex-ctl{display: grid;gap: 10px;margin-top: 16px}.ag-root.ag-root .ag-ex-head{display: grid;grid-template-columns: minmax(0,40%) minmax(0,1fr) 66px;gap: 14px;align-items: end;margin-top: 18px;padding-bottom: 6px;border-bottom: 1px solid var(--ag-grid)}.ag-root.ag-root .ag-ex-legend{display: flex;flex-wrap: wrap;gap: 4px 12px;font-size: 12px;color: var(--ag-ink-2)}.ag-root.ag-root .ag-ex-legend span{display: inline-flex;align-items: center;gap: 5px}.ag-root.ag-root .ag-ex-axis{position: relative;height: 16px;font-size: 11px;color: var(--ag-ink-3);font-variant-numeric: tabular-nums}.ag-root.ag-root .ag-ex-axis span{position: absolute;bottom: 0;transform: translateX(-50%);white-space: nowrap}.ag-root.ag-root .ag-ex-axis span:first-child{transform: none}.ag-root.ag-root .ag-ex-axis span:last-child{transform: translateX(-100%)}.ag-root.ag-root .ag-ex-pcth{font-size: 11px;color: var(--ag-ink-3);text-align: right;line-height: 1.2}.ag-root.ag-root .ag-ex-list{list-style: none}.ag-root.ag-root .ag-ex-row{border-bottom: 1px solid var(--ag-grid)}.ag-root.ag-root .ag-ex-btn{display: grid;grid-template-columns: minmax(0,40%) minmax(0,1fr) 66px;gap: 14px;align-items: center;width: 100%;padding: 7px 0;text-align: left;transition: background .15s}.ag-root.ag-root .ag-ex-btn:hover{background: color-mix(in srgb,var(--ag-card) 70%,transparent)}.ag-root.ag-root .ag-ex-names{display: grid;gap: 1px;font-size: 13px;line-height: 1.3;min-width: 0}.ag-root.ag-root .ag-ex-names span{display: flex;gap: 6px;align-items: baseline;min-width: 0}.ag-root.ag-root .ag-ex-names span>i{overflow: hidden;text-overflow: ellipsis;white-space: nowrap;font-style: normal}.ag-root.ag-root .ag-ex-names .ag-sw{width: 8px;height: 8px;transform: translateY(-1px)}.ag-root.ag-root .ag-ex-names .is-f{color: var(--ag-ink-2)}.ag-root.ag-root .ag-ex-track{position: relative;height: 22px}.ag-root.ag-root .ag-ex-track::before{content: "";position: absolute;left: 0;right: 0;top: 50%;height: 1px;background: var(--ag-track)}.ag-root.ag-root .ag-ex-gl{position: absolute;top: 0;bottom: 0;width: 1px;background: var(--ag-track)}.ag-root.ag-root .ag-ex-bar{position: absolute;top: 50%;height: 3px;margin-top: -1.5px;border-radius: 2px;opacity: .45;transition: left .7s var(--ag-ease),width .7s var(--ag-ease)}.ag-root.ag-root .ag-ex-bar.is-m{background: var(--ag-m)}.ag-root.ag-root .ag-ex-bar.is-f{background: var(--ag-f)}.ag-root.ag-root .ag-ex-dot{position: absolute;top: 50%;width: 13px;height: 13px;margin: -6.5px 0 0 -6.5px;border-radius: 50%;box-shadow: 0 0 0 2px var(--ag-knock);transition: left .7s var(--ag-ease)}.ag-root.ag-root .ag-ex-dot.is-m{background: var(--ag-m)}.ag-root.ag-root .ag-ex-dot.is-f{background: var(--ag-f)}.ag-root.ag-root .ag-ex-pct{text-align: right;font-size: 14px;font-weight: 650;font-variant-numeric: tabular-nums;display: flex;justify-content: flex-end;align-items: center;gap: 5px}.ag-root.ag-root .ag-ex-pct .ag-sw{width: 7px;height: 7px}.ag-root.ag-root .ag-ex-detail{padding: 4px 0 14px;display: grid;gap: 8px;animation: ag-in .35s var(--ag-ease) both}.ag-root.ag-root .ag-cpt{color: var(--ag-ink-3);font-size: 12px;font-weight: 400;margin-left: 6px;white-space: nowrap;font-variant-numeric: tabular-nums}.ag-root.ag-root .ag-flag{font-size: 12px;color: var(--ag-ink-2);border-left: 2px solid var(--ag-grid);padding: 2px 0 2px 10px}.ag-root.ag-root .ag-ex-none{padding: 24px 0;text-align: center;color: var(--ag-ink-3);font-size: 14px}.ag-root.ag-root .ag-ex-row.is-enter{animation: ag-in .45s var(--ag-ease) both}@container ag (max-width: 540px){.ag-root.ag-root .ag-ex-head{grid-template-columns: minmax(0,1fr) 60px}.ag-root.ag-root .ag-ex-head .ag-ex-legend{grid-column: 1 / -1}.ag-root.ag-root .ag-ex-btn{grid-template-columns: minmax(0,1fr) 60px;gap: 4px 12px;padding: 10px 0}.ag-root.ag-root .ag-ex-names{grid-column: 1 / -1}}.ag-root.ag-root .ag-cpt{color: var(--ag-ink-3);font-size: 12px;font-weight: 400;font-style: normal;margin-left: 6px;white-space: nowrap;font-variant-numeric: tabular-nums;letter-spacing: 0}.ag-root.ag-root .ag-cpt-wrap{position: relative;display: inline-block}.ag-root.ag-root .ag-cpt-btn{padding: 2px 0;margin-left: 6px;line-height: 1.2;cursor: help;text-decoration: underline dotted;text-decoration-color: var(--ag-ink-3);text-underline-offset: 3px}.ag-root.ag-root .ag-cpt-btn:hover,.ag-root.ag-root .ag-cpt-wrap.is-open .ag-cpt-btn{color: var(--ag-ink);text-decoration-color: currentColor}.ag-root.ag-root .ag-tip{position: absolute;left: 0;bottom: calc(100% + 6px);z-index: 5;width: max-content;max-width: min(260px,72vw);padding: 8px 10px;border-radius: 6px;background: var(--ag-ink);color: var(--ag-knock);font-size: 13px;font-weight: 500;line-height: 1.35;white-space: normal;text-align: left;box-shadow: 0 4px 14px color-mix(in srgb,var(--ag-ink) 18%,transparent);display: none;pointer-events: none}.ag-root.ag-root .ag-cpt-wrap.is-open .ag-tip{display: block;animation: ag-fade .15s linear both}@keyframes ag-fade{from{opacity: 0}to{opacity: 1}}.ag-root.ag-root .ag-tip-k{display: block;margin-bottom: 2px;font-size: 11px;font-weight: 500;opacity: .7}.ag-root.ag-root .ag-pair{display: grid;gap: 4px}.ag-root.ag-root .ag-pair-row{display: flex;justify-content: space-between;align-items: baseline;gap: 12px}.ag-root.ag-root .ag-pair-name{font-size: 15px;font-weight: 500;line-height: 1.3;min-width: 0}.ag-root.ag-root .ag-pair-val{font-size: 15px;font-weight: 600;white-space: nowrap}.ag-root.ag-root .ag-find{position: relative}.ag-root.ag-root .ag-find svg{position: absolute;left: 0;top: 50%;width: 15px;height: 15px;margin-top: -7.5px;color: var(--ag-ink-3);pointer-events: none}.ag-root.ag-root .ag-find input{display: block;width: 100%;min-height: 40px;padding: 8px 0 8px 24px;border-radius: 0;background: transparent;border-bottom: 1px solid var(--ag-grid);font-size: 16px;color: var(--ag-ink);-webkit-appearance: none;appearance: none;transition: border-color .2s}.ag-root.ag-root .ag-find input:hover{border-bottom-color: var(--ag-ink-3)}.ag-root.ag-root .ag-find input:focus-visible{outline: none;border-bottom-color: var(--ag-ink)}.ag-root.ag-root .ag-find input::placeholder{color: var(--ag-ink-3);opacity: 1}.ag-root.ag-root .ag-find input::-webkit-search-cancel-button{-webkit-appearance: none}.ag-root.ag-root.ag-root .ag-q-btn{min-height: 42px;padding: 10px 20px;border-radius: var(--ag-r);background: transparent;color: var(--ag-ink);box-shadow: inset 0 0 0 1px var(--ag-ink);font-size: 15px;font-weight: 500;transition: background-color .2s,color .2s,transform .15s}.ag-root.ag-root.ag-root .ag-q-btn:hover{background: var(--ag-ink);color: var(--ag-knock)}.ag-root.ag-root .ag-pt-find{margin-top: 10px;max-width: 340px}.ag-root.ag-root .ag-pt-hits{list-style: none;margin-top: 6px;display: grid;gap: 0}.ag-root.ag-root .ag-pt-hits:empty{display: none}.ag-root.ag-root .ag-pt-hit{display: block;width: 100%;padding: 6px 0;text-align: left;font-size: 14px;font-weight: 500;line-height: 1.3}.ag-root.ag-root .ag-pt-hit:hover{text-decoration: underline;text-underline-offset: 3px;text-decoration-color: var(--ag-grid)}.ag-root.ag-root .ag-pt-none{padding: 6px 0;font-size: 13px;color: var(--ag-ink-3)}.ag-root.ag-root .ag-pt-finding .ag-pt-dot:not(.is-hit):not(.is-on){opacity: .15}.ag-root.ag-root .ag-pt-dot.is-hit{stroke: var(--ag-ink);stroke-width: 2px}.ag-root.ag-root .ag-pt-cap{margin-top: 6px}.ag-root.ag-root .ag-pt-cap>span{display: inline-flex;align-items: baseline;gap: 5px;font-weight: 500;white-space: nowrap}@container ag (max-width: 340px){.ag-root.ag-root .ag-pt-cap{font-size: 11px;gap: 6px}.ag-root.ag-root .ag-pt-cap>span{gap: 3px}}.ag-root.ag-root .ag-arrow{font-weight: 400;color: var(--ag-ink-3)}.ag-root.ag-root .ag-pt-card{margin-top: 12px;min-height: 0;padding-top: 12px;border-top: 1px solid var(--ag-grid);display: grid;gap: 8px}.ag-root.ag-root .ag-pt-empty{font-size: 14px;color: var(--ag-ink-3)}.ag-root.ag-root .ag-ex-ctl{display: flex;flex-wrap: wrap;gap: 12px 32px;align-items: end;margin-top: 16px}.ag-root.ag-root .ag-ex-ctl .ag-find{flex: 1 1 220px}.ag-root.ag-root .ag-ex-sels{display: flex;gap: 20px;flex-wrap: wrap}.ag-root.ag-root .ag-mini{display: grid;gap: 2px}.ag-root.ag-root .ag-mini-k{font-size: 12px;color: var(--ag-ink-3)}.ag-root.ag-root .ag-mini-v{position: relative;display: block}.ag-root.ag-root .ag-mini-v::after{content: "";position: absolute;right: 2px;top: 50%;width: 6px;height: 6px;margin-top: -5px;border-right: 1.5px solid var(--ag-ink-2);border-bottom: 1.5px solid var(--ag-ink-2);transform: rotate(45deg);pointer-events: none}.ag-root.ag-root .ag-mini-sel{display: block;min-height: 40px;padding: 8px 20px 8px 0;background: transparent;border-bottom: 1px solid var(--ag-grid);font-size: 15px;font-weight: 500;color: var(--ag-ink);-webkit-appearance: none;appearance: none;cursor: pointer;border-radius: 0}.ag-root.ag-root .ag-mini-sel:hover{border-bottom-color: var(--ag-ink-3)}.ag-root.ag-root .ag-mini-sel:focus-visible{outline: none;border-bottom-color: var(--ag-ink)}.ag-root.ag-root .ag-ex-head{margin-top: 24px}.ag-root.ag-root .ag-ex-names .ag-ex-n{overflow: hidden;text-overflow: ellipsis;white-space: nowrap;font-style: normal;min-width: 0}.ag-root.ag-root .ag-ex-names .ag-cpt{margin-left: 0;flex: none;font-size: 11px}.ag-root.ag-root .ag-ex-names .is-f{color: inherit}.ag-root.ag-root .ag-ex-pct{font-weight: 600}.ag-root.ag-root .ag-ex-detail{padding: 4px 0 12px;gap: 8px}.ag-root.ag-root .ag-ex-detail .ag-pair-name,.ag-root.ag-root .ag-ex-detail .ag-pair-val{font-size: 15px}.ag-root.ag-root .ag-ex-pager{display: flex;justify-content: space-between;align-items: center;gap: 12px;margin-top: 8px}.ag-root.ag-root .ag-ex-count{font-size: 13px;color: var(--ag-ink-3);font-variant-numeric: tabular-nums}.ag-root.ag-root .ag-ex-pbtns{display: flex;gap: 8px}.ag-root.ag-root .ag-icon-btn[disabled]{opacity: .3;cursor: default}@container ag (max-width: 480px){.ag-root.ag-root .ag-ex-sels{width: 100%}.ag-root.ag-root .ag-ex-sels .ag-mini{flex: 1 1 0;min-width: 0}.ag-root.ag-root .ag-mini-sel{width: 100%}}.ag-root.ag-root .ag-stat{display: block;text-wrap: pretty}.ag-root.ag-root .ag-stat-num{font-family: var(--ag-head);font-weight: 600;font-size: 30px;line-height: 1.05;letter-spacing: -0.025em;font-variant-numeric: tabular-nums;color: var(--ag-ink);margin-right: 8px}.ag-root.ag-root .ag-stat-cap{font-size: 15px;font-weight: 500;line-height: 1.35;color: var(--ag-ink-2)}.ag-root.ag-root .ag-stat-cap span{white-space: nowrap}.ag-root.ag-root .ag-stat-num.ag-ink-m{color: var(--ag-m-ink)}.ag-root.ag-root .ag-stat-num.ag-ink-f{color: var(--ag-f-ink)}.ag-root.ag-root .ag-stat.is-small .ag-stat-num{font-size: 24px}.ag-root.ag-root .ag-stat.is-small .ag-stat-cap{font-size: 14px}.ag-root.ag-root .ag-mu-ctl{display: flex;gap: 6px;align-items: stretch;margin-top: 14px}.ag-root.ag-root .ag-select-wrap{position: relative;flex: 1 1 auto;min-width: 0}.ag-root.ag-root .ag-select{display: block;-webkit-appearance: none;appearance: none;cursor: pointer;font-size: 16px}.ag-root.ag-root .ag-select-wrap .ag-select{position: absolute;inset: 0;width: 100%;height: 100%;opacity: 0;min-height: 0}.ag-root.ag-root .ag-select-face{display: flex;align-items: center;min-height: 40px;height: 100%;padding: 6px 34px 6px 12px;border-radius: var(--ag-r);background: transparent;box-shadow: inset 0 0 0 1px var(--ag-grid);transition: box-shadow .2s;font-size: 15px;font-weight: 500;line-height: 1.3}.ag-root.ag-root .ag-select-wrap:hover .ag-select-face{box-shadow: inset 0 0 0 1px var(--ag-ink-3)}.ag-root.ag-root .ag-select-wrap:has(.ag-select:focus-visible) .ag-select-face{outline: 2px solid var(--ag-focus);outline-offset: 2px}.ag-root.ag-root .ag-select-wrap::after{content: "";position: absolute;right: 14px;top: 50%;width: 6px;height: 6px;margin-top: -5px;border-right: 1.5px solid var(--ag-ink-2);border-bottom: 1.5px solid var(--ag-ink-2);transform: rotate(45deg);pointer-events: none}.ag-root.ag-root .ag-mu-ctl .ag-icon-btn{width: 40px;height: auto;min-height: 40px}.ag-root.ag-root .ag-mu-ctl .ag-icon-btn svg{width: 14px;height: 14px}.ag-root.ag-root .ag-mu-sum{margin-top: 14px}@container ag (max-width: 480px){.ag-root.ag-root .ag-mu-ctl{gap: 4px}.ag-root.ag-root .ag-mu-ctl .ag-icon-btn{width: 32px}.ag-root.ag-root .ag-select-face{font-size: 14px;padding-left: 10px;padding-right: 28px}.ag-root.ag-root .ag-select-wrap::after{right: 11px}}.ag-root.ag-root .ag-mu-strip-wrap{margin-top: 12px}.ag-root.ag-root .ag-mu-strip-lab{font-size: 13px;color: var(--ag-ink-3)}.ag-root.ag-root .ag-mu-strip{position: relative;height: 48px;margin-top: 2px}.ag-root.ag-root .ag-mu-strip::before{content: "";position: absolute;left: 0;right: 0;top: 31px;height: 1px;background: var(--ag-grid)}.ag-root.ag-root .ag-mu-zero{position: absolute;top: 15px;bottom: 0;width: 0;border-left: 1.5px dashed var(--ag-ink-2);margin-left: -0.75px}.ag-root.ag-root .ag-mu-zero-lab{position: absolute;top: 0;transform: translateX(-50%);font-size: 11px;font-weight: 500;line-height: 1;color: var(--ag-ink-2);white-space: nowrap}.ag-root.ag-root .ag-mu-tick{position: absolute;top: 23px;width: 2px;height: 16px;margin-left: -1px;padding: 0;border-radius: 1px;opacity: .45;transition: height .25s var(--ag-ease),top .25s var(--ag-ease),width .25s var(--ag-ease),opacity .2s}.ag-root.ag-root .ag-mu-tick.is-m{background: var(--ag-m)}.ag-root.ag-root .ag-mu-tick.is-f{background: var(--ag-f)}.ag-root.ag-root .ag-mu-tick::after{content: "";position: absolute;inset: -8px -3px}.ag-root.ag-root .ag-mu-tick:hover{opacity: .9}.ag-root.ag-root .ag-mu-tick.is-on{opacity: 1;top: 16px;height: 30px;width: 4px;margin-left: -2px;box-shadow: 0 0 0 2px var(--ag-knock);z-index: 2}.ag-root.ag-root .ag-mu-pair{margin-top: 14px;padding-top: 12px;border-top: 1px solid var(--ag-grid);display: grid;gap: 10px}.ag-root.ag-root .ag-mu-side{display: grid;gap: 4px}.ag-root.ag-root .ag-mu-line{display: flex;justify-content: space-between;align-items: baseline;gap: 12px}.ag-root.ag-root .ag-mu-name{font-size: 15px;font-weight: 500;line-height: 1.3;min-width: 0}.ag-root.ag-root .ag-mu-val{font-size: 15px;font-weight: 600;white-space: nowrap}.ag-root.ag-root .ag-mu-pair .ag-bar-track{height: 12px}.ag-root.ag-root .ag-q-prog{display: block;margin: 0 0 8px;font-size: 12px;font-weight: 500;color: var(--ag-ink-3)}.ag-root.ag-root .ag-q-lab{display: inline-flex;align-items: baseline;flex-wrap: wrap;font-size: 15px;font-weight: 500;line-height: 1.3}.ag-root.ag-root .ag-q-row-hd b{font-weight: 600}.ag-root.ag-root .ag-q-foot{display: flex;flex-wrap: wrap;align-items: center;justify-content: space-between;gap: 10px 16px;margin-top: 6px}.ag-root.ag-root .ag-q-foot.is-result{margin-top: 26px}.ag-root.ag-root .ag-q-endline{font-size: 17px;font-weight: 500;line-height: 1.35;color: var(--ag-ink);text-wrap: pretty}.ag-root.ag-root .ag-q-endnum{font-family: var(--ag-head);font-weight: 600;font-size: 30px;letter-spacing: -0.025em}.ag-root.ag-root .ag-q-hint{display: flex;flex-wrap: wrap;align-items: center;gap: 4px;font-size: 12px;color: var(--ag-ink-3)}.ag-root.ag-root .ag-q-ring{display: inline-block;width: 10px;height: 10px;border-radius: 50%;box-shadow: inset 0 0 0 2px var(--ag-ink-2)}.ag-root.ag-root .ag-q-cta{display: flex;flex-wrap: wrap;align-items: center;gap: 10px 18px;margin-top: 12px}.ag-root.ag-root .ag-q-link{font-size: 15px;font-weight: 500;color: var(--ag-ink);text-decoration: underline;text-decoration-color: var(--ag-grid);text-underline-offset: 4px}.ag-root.ag-root .ag-q-link:hover{text-decoration-color: currentColor}.ag-root.ag-root{--ag-ui: "Geist","Helvetica Neue",Arial,sans-serif;--ag-head-weight: 600;--ag-head-track: -0.02em;--ag-r: 8px}.ag-root.ag-root[data-ag-mode="light"]{--ag-card: #f3f3f6;--ag-grid: #e2e2e8;--ag-track: #ececf1}@media (prefers-color-scheme: light){.ag-root.ag-root[data-ag-mode="auto"]{--ag-card: #f3f3f6;--ag-grid: #e2e2e8;--ag-track: #ececf1}}.ag-root.ag-root[data-ag-style="editorial"]{--ag-head-weight: 700}.ag-root.ag-root[data-ag-style="feature"]{--ag-head-weight: 700}.ag-root.ag-root.ag-root{font-family: var(--ag-ui)}.ag-root.ag-root.ag-root .ag-hed{letter-spacing: var(--ag-head-track)}.ag-root.ag-root.ag-root .ag-mu-name{font-weight: 500}.ag-root.ag-root.ag-root .ag-q-row-hd b,.ag-root.ag-root.ag-root .ag-mu-val,.ag-root.ag-root.ag-root .ag-pair-val,.ag-root.ag-root.ag-root .ag-ex-pct{font-weight: 600}.ag-root.ag-root.ag-root .ag-dek{text-wrap: pretty}.ag-root.ag-root.ag-root .ag-btn,.ag-root.ag-root.ag-root .ag-icon-btn,.ag-root.ag-root.ag-root .ag-select-face{border-radius: var(--ag-r)}.ag-root.ag-root.ag-root .ag-bar{border-radius: 0 2px 2px 0}.ag-root.ag-root.ag-root .ag-btn{font-weight: 600}.ag-root.ag-root.ag-root .ag-btn{transition: transform .15s var(--ag-ease),background-color .2s}.ag-root.ag-root.ag-root .ag-btn:hover{opacity: 1;background: color-mix(in srgb,var(--ag-ink) 82%,var(--ag-knock))}.ag-root.ag-root.ag-root .ag-btn:active,.ag-root.ag-root.ag-root .ag-icon-btn:active{transform: translateY(1px)}.ag-root.ag-root.ag-root .ag-icon-btn{transition: box-shadow .2s,background-color .2s,transform .15s}.ag-root.ag-root.ag-root .ag-icon-btn:hover{background: var(--ag-card)}.ag-root.ag-root.ag-root .ag-btn-ghost{background: none;box-shadow: none;padding-left: 4px;padding-right: 4px;font-weight: 500;text-decoration: underline;text-decoration-color: var(--ag-grid);text-underline-offset: 4px}.ag-root.ag-root.ag-root .ag-btn-ghost:hover{background: none;text-decoration-color: currentColor}.ag-root.ag-root.ag-root .ag-q-prog span{margin-left: 0;font-weight: 500}.ag-root.ag-root.ag-root .ag-q-mini{background: none;height: 12px}.ag-root.ag-root.ag-root .ag-q-mini::before{content: "";position: absolute;left: 0;right: 0;top: 50%;height: 1px;background: var(--ag-grid)}.ag-root.ag-root.ag-root .ag-q-mini i{top: 0}.ag-root.ag-root.ag-root .ag-range::-webkit-slider-thumb{box-shadow: 0 1px 3px color-mix(in srgb,var(--ag-ink) 22%,transparent)}.ag-root.ag-root.ag-root .ag-range::-moz-range-thumb{box-shadow: 0 1px 3px color-mix(in srgb,var(--ag-ink) 22%,transparent)}@media (prefers-reduced-motion: reduce){.ag-root.ag-root *,.ag-root.ag-root *::before,.ag-root.ag-root *::after{transition: none !important;animation: none !important}} (window.AG = window.AG || {}).data = [{"id":1,"cat":"Biopsies & lesions","mName":"Biopsy penis","mCpt":"54100","m":121.32,"fName":"Biopsy vagina","fCpt":"57100","f":65.74,"pct":59.42},{"id":2,"cat":"Biopsies & lesions","mName":"Remove penis lesion","mCpt":"54110","m":624.88,"fName":"Remove vaginal lesion","fCpt":"57135","f":191.8,"pct":106.06},{"id":3,"cat":"Biopsies & lesions","mName":"Biopsy scrotum","mCpt":"55120","m":356.83,"fName":"Biopsy vulva","fCpt":"56605","f":59.98,"pct":142.44},{"id":4,"cat":"Biopsies & lesions","mName":"Biopsy prostate","mCpt":"55705","m":265.34,"fName":"Biopsy endometrium","fCpt":"58110","f":40.33,"pct":147.22},{"id":5,"cat":"Biopsies & lesions","mName":"Removal of scrotal lesion","mCpt":"55120","m":356.83,"fName":"Removal of vulvar lesion","fCpt":"56440","f":184.01,"pct":63.91},{"id":6,"cat":"Biopsies & lesions","mName":"Excision penile lesion","mCpt":"54115","m":428.0,"fName":"Excision vaginal lesion","fCpt":"57100","f":65.74,"pct":146.74},{"id":7,"cat":"Biopsies & lesions","mName":"Laser penis","mCpt":"54057","m":98.95,"fName":"Laser vulva","fCpt":"56501","f":136.57,"pct":-31.95},{"id":8,"cat":"Genitals & other","mName":"Distal hypospadias repair","mCpt":"54328","m":934.61,"fName":"Revision of cervix","fCpt":"57700","f":365.3,"pct":87.59},{"id":9,"cat":"Biopsies & lesions","mName":"Destruction penile lesions simple","mCpt":"54050","m":108.1,"fName":"Destruction vulvar lesions simple","fCpt":"56501","f":136.57,"pct":-23.27},{"id":10,"cat":"Genitals & other","mName":"Resection of scrotum","mCpt":"55150","m":496.11,"fName":"Simple complete vulvectomy","fCpt":"56620","f":598.11,"pct":-18.64},{"id":11,"cat":"Genitals & other","mName":"Revision of scrotum","mCpt":"55180","m":690.96,"fName":"Extensive vulva surgery (partial radical)","fCpt":"56630","f":976.63,"pct":-34.26},{"id":12,"cat":"Genitals & other","mName":"Drain prostate abscess","mCpt":"52700","m":443.92,"fName":"Drain ovarian abscess open","fCpt":"58820","f":347.01,"pct":24.51},{"id":13,"cat":"Genitals & other","mName":"Excision hydrocele","mCpt":"55040","m":339.89,"fName":"Excision Bartholins gland","fCpt":"56740","f":321.25,"pct":5.64},{"id":14,"cat":"Genitals & other","mName":"Penis/scrotoplasty","mCpt":"54360","m":721.8,"fName":"Perineoplasty","fCpt":"56810","f":276.86,"pct":89.11},{"id":15,"cat":"Genitals & other","mName":"Excision of sperm cord lesion","mCpt":"55520","m":465.95,"fName":"Excision of Bartholins","fCpt":"56740","f":321.25,"pct":36.76},{"id":16,"cat":"Prostate & uterus","mName":"Interstitial prostate","mCpt":"55875","m":780.08,"fName":"Interstitial cervix","fCpt":"55920","f":463.24,"pct":50.97},{"id":17,"cat":"Urinary tract","mName":"Insert tandem male","mCpt":"53444","m":792.62,"fName":"Insert tandem female","fCpt":"57155","f":286.01,"pct":93.94},{"id":18,"cat":"Urinary tract","mName":"Reconstruction of male urethra","mCpt":"54352","m":1408.35,"fName":"Reconstruction of female urethra","fCpt":"53430","f":975.27,"pct":36.34},{"id":19,"cat":"Urinary tract","mName":"Revise remove sling male","mCpt":"53448","m":1271.79,"fName":"Revise remove sling female","fCpt":"57287","f":754.33,"pct":51.08},{"id":20,"cat":"Urinary tract","mName":"Male sling insert uro/neck sph","mCpt":"53445","m":757.72,"fName":"Female insert mesh pelvic floor add on","fCpt":"57267","f":252.12,"pct":100.13},{"id":21,"cat":"Urinary tract","mName":"Reconstruct urethra male","mCpt":"53415","m":1126.07,"fName":"Revise urethra female","fCpt":"57220","f":353.11,"pct":104.51},{"id":22,"cat":"Urinary tract","mName":"Total urethrectomy male","mCpt":"53215","m":926.14,"fName":"Total urethrectomy female","fCpt":"53210","f":776.7,"pct":17.55},{"id":23,"cat":"Genitals & other","mName":"Partial removal of penis","mCpt":"54120","m":633.01,"fName":"Partial vaginectomy","fCpt":"57106","f":546.94,"pct":14.59},{"id":24,"cat":"Genitals & other","mName":"Removal of penis","mCpt":"54125","m":825.49,"fName":"Complete vaginectomy","fCpt":"57110","f":919.02,"pct":-10.72},{"id":25,"cat":"Genitals & other","mName":"Exploratory laparotomy male","mCpt":"49000","m":778.05,"fName":"Exploratory laparotomy female","fCpt":"58960","f":1009.84,"pct":-25.93},{"id":26,"cat":"Testes & ovaries","mName":"Excise lesion testis","mCpt":"54512","m":538.47,"fName":"Ovarian cystectomy","fCpt":"58925","f":781.44,"pct":-36.82},{"id":27,"cat":"Testes & ovaries","mName":"Removal of testis","mCpt":"54530","m":511.02,"fName":"Oophorectomy","fCpt":"58940","f":567.27,"pct":-10.43},{"id":28,"cat":"Testes & ovaries","mName":"Suspension/relocation of testes","mCpt":"54680","m":787.54,"fName":"Ovarian transposition","fCpt":"58825","f":722.81,"pct":8.57},{"id":29,"cat":"Testes & ovaries","mName":"Laparoscopic orchiectomy","mCpt":"54690","m":655.38,"fName":"Laparoscopic BSO","fCpt":"58661","f":662.83,"pct":-1.13},{"id":30,"cat":"Testes & ovaries","mName":"Laparoscopic orchiopexy","mCpt":"54692","m":755.01,"fName":"Laparoscopic ovarian transposition","fCpt":"58825","f":722.81,"pct":4.36},{"id":31,"cat":"Testes & ovaries","mName":"Removal of epididymis","mCpt":"54861","m":569.64,"fName":"Removal of fallopian tubes","fCpt":"58700","f":814.31,"pct":-35.36},{"id":32,"cat":"Testes & ovaries","mName":"Vas deferens vasectomy","mCpt":"55250","m":230.43,"fName":"Essure hysteroscopy","fCpt":"58565","f":466.97,"pct":-67.83},{"id":33,"cat":"Testes & ovaries","mName":"Orchiectomy with staging","mCpt":"54530 + 38770","m":1320.25,"fName":"Oophorectomy with staging","fCpt":"58950","f":1169.79,"pct":12.08},{"id":34,"cat":"Testes & ovaries","mName":"Extensive testis surgery","mCpt":"54535","m":744.5,"fName":"Removal of ovaries extensive","fCpt":"58943","f":1188.42,"pct":-45.93},{"id":35,"cat":"Testes & ovaries","mName":"Reduce testis torsion","mCpt":"54600","m":454.43,"fName":"Reduce ovarian torsion","fCpt":"58920","f":728.24,"pct":-46.3},{"id":36,"cat":"Genitals & other","mName":"Remove penis and LN","mCpt":"54135","m":1503.24,"fName":"Vaginectomy and LN","fCpt":"57109","f":1751.97,"pct":-15.28},{"id":37,"cat":"Testes & ovaries","mName":"Revise vas deferens/spermatic ducts","mCpt":"54901","m":1056.94,"fName":"Revise fallopian tubes","fCpt":"58740","f":917.67,"pct":14.11},{"id":38,"cat":"Prostate & uterus","mName":"TURP","mCpt":"52601","m":727.22,"fName":"Hysteroscopic myomectomy","fCpt":"58561","f":360.9,"pct":67.33},{"id":39,"cat":"Prostate & uterus","mName":"Remove prostate","mCpt":"55801","m":1093.2,"fName":"Hysterectomy","fCpt":"58150","f":1028.82,"pct":6.07},{"id":40,"cat":"Prostate & uterus","mName":"Extensive prostate surgery","mCpt":"55845","m":1355.49,"fName":"Extensive TAH","fCpt":"58285","f":1440.21,"pct":-6.06},{"id":41,"cat":"Prostate & uterus","mName":"Subtotal prostatectomy","mCpt":"55821","m":837.69,"fName":"Abdominal trachelectomy","fCpt":"57530","f":381.57,"pct":74.82},{"id":42,"cat":"Prostate & uterus","mName":"Prostatectomy","mCpt":"55840","m":1165.38,"fName":"TAH","fCpt":"58150","f":1028.82,"pct":12.45},{"id":43,"cat":"Prostate & uterus","mName":"Prostatectomy LND","mCpt":"55845","m":1355.49,"fName":"TAH BSO LND","fCpt":"58200","f":1362.6,"pct":-0.52},{"id":44,"cat":"Prostate & uterus","mName":"Perineal approach prostatectomy","mCpt":"55815","m":1751.97,"fName":"Radical vaginal hysterectomy","fCpt":"58285","f":1440.21,"pct":19.53},{"id":45,"cat":"Prostate & uterus","mName":"Radical prostatectomy perineal approach","mCpt":"55810","m":1301.27,"fName":"Radical vaginal hysterectomy","fCpt":"58285","f":1440.21,"pct":-10.14},{"id":46,"cat":"Urinary tract","mName":"Removal of bladder and LN","mCpt":"51575","m":1809.92,"fName":"Anterior exenteration (no code for anterior only)","fCpt":"58240","f":2967.84,"pct":-48.47},{"id":47,"cat":"Urinary tract","mName":"Cystectomy complete with conduit","mCpt":"51596","m":2344.32,"fName":"Pelvic exenteration","fCpt":"58240","f":2967.84,"pct":-23.48},{"id":48,"cat":"Biopsies & lesions","mName":"Biopsy penis","mCpt":"54100","m":121.32,"fName":"Biopsy vulva","fCpt":"56605","f":59.98,"pct":67.67},{"id":49,"cat":"Biopsies & lesions","mName":"Biopsy prostate","mCpt":"55700","m":129.11,"fName":"Biopsy endometrium","fCpt":"58100","f":63.71,"pct":67.84},{"id":50,"cat":"Biopsies & lesions","mName":"Destruction penile lesions extensive","mCpt":"54065","m":172.83,"fName":"Destruction vulvar lesions extensive","fCpt":"56515","f":216.88,"pct":-22.61},{"id":51,"cat":"Prostate & uterus","mName":"TURP","mCpt":"52601","m":727.22,"fName":"Hysteroscopy","fCpt":"58558","f":233.48,"pct":102.79},{"id":52,"cat":"Genitals & other","mName":"Drain prostate abscess","mCpt":"55720","m":453.07,"fName":"Drain ovarian abscess","fCpt":"58820","f":347.01,"pct":26.51},{"id":53,"cat":"Genitals & other","mName":"Excision varicocele","mCpt":"55530","m":354.12,"fName":"Excision Bartholins","fCpt":"56740","f":321.25,"pct":9.73},{"id":54,"cat":"Genitals & other","mName":"Excision hydrocele","mCpt":"55500","m":395.8,"fName":"Excision Bartholins","fCpt":"56740","f":321.25,"pct":20.79},{"id":55,"cat":"Genitals & other","mName":"Scrotoplasty","mCpt":"55175","m":367.68,"fName":"Perineoplasty","fCpt":"56810","f":276.86,"pct":28.18}]; window.AG.copy = {"shared":{"method":"Percent difference is the gap between the two payments divided by their average. Because neither the male nor female procedure serves as the baseline to compare against, the average is used, yielding a single figure for each pair. Negative values mean the procedure on women pays more.","source":"Source: Penn et al., Journal of Women\u2019s Health (2025), Table 3.","credit":"Graphic: Anusha Subramanian/Vox","logoLabel":"Vox","aboutButton":"About the data","about1":"Amounts are Medicare\u2019s 2023 national reimbursement rates when the procedure is done in a hospital or surgical center.","about2":"Researchers matched each procedure on male anatomy with a comparable one on female anatomy. Some procedures appear in more than one pair. CPT codes are the billing codes listed in the study. Look any code up in Medicare\u2019s Physician Fee Schedule tool.","difference":"pay difference","srHigherMale":"higher for the male procedure","srHigherFemale":"higher for the female procedure","paysMoreFemale":"Pays more for female","paysMoreMale":"Pays more for male","noMatches":"No matching procedures","cptPrefix":"CPT"},"pattern":{"hed":"Medicare often reimburses more for procedures on men","dek":"Percent difference in 2023 Medicare reimbursement rates among pairs of comparable male and female procedures. Each dot is one pair.","findPlaceholder":"Find a procedure or code","emptyCard":"Tap or hover over a dot to see the pair","moreHits":"and {n} more, highlighted on the chart"},"quiz":{"hed":"Can you guess what Medicare pays?","dek":"Drag the slider to estimate what Medicare pays for the second procedure.","round":"Round {i} of {n}","submit":"Submit","next":"Next pair","seeResult":"See the result","yourGuess":"Your guess {amount}","endLede":"{n} of {total}","endText":"matched pairs pay the surgeon more for the procedure on male anatomy.","legendMale":"Male","legendFemale":"Female","legendGuess":"Your guess","playAgain":"Play again with new pairs","seeAll":"See all 55 pairs"},"matchup":{"hed":"Similar procedures, different pay","dek":"What Medicare paid a surgeon for matched procedures in 2023","stripLabel":"Where this pair falls among all 55","noDifference":"No difference"},"explorer":{"hed":"Every matched pair, side by side","dek":"What Medicare paid a surgeon for each procedure in 2023","searchPlaceholder":"Search a procedure or code","legendMale":"Male procedure","legendFemale":"Female procedure","pctHeader":"% diff.","count":"{from}\u2013{to} of {n} pairs","countFiltered":"{from}\u2013{to} of {n} matching pairs","sortGap":"Sorted by largest difference","sortStudy":"Sorted in the study\u2019s order","sortPay":"Sorted by highest payment","essureFlag":"Essure was taken off the U.S. market in 2018."}}; (function () { var AG = (window.AG = window.AG || {}); AG.components = AG.components || {}; if (AG.__core) return; AG.__core = true; AG.CATEGORIES = ["Biopsies & lesions", "Testes & ovaries", "Prostate & uterus", "Urinary tract", "Genitals & other"]; var P = function (d) { return ''; }; AG.icons = { prev: P("M168.49,199.51a12,12,0,0,1-17,17l-80-80a12,12,0,0,1,0-17l80-80a12,12,0,0,1,17,17L97,128Z"), next: P("M184.49,136.49l-80,80a12,12,0,0,1-17-17L159,128,87.51,56.49a12,12,0,1,1,17-17l80,80A12,12,0,0,1,184.49,136.49Z"), shuffle: P("M240.49,175.51a12,12,0,0,1,0,17l-24,24a12,12,0,0,1-17-17L203,196h-2.09a76.17,76.17,0,0,1-61.85-31.83L97.38,105.78A52.1,52.1,0,0,0,55.06,84H32a12,12,0,0,1,0-24H55.06a76.17,76.17,0,0,1,61.85,31.83l41.71,58.39A52.1,52.1,0,0,0,200.94,172H203l-3.52-3.51a12,12,0,0,1,17-17Zm-95.62-72.62a12,12,0,0,0,16.93-1.13A52,52,0,0,1,200.94,84H203l-3.52,3.51a12,12,0,0,0,17,17l24-24a12,12,0,0,0,0-17l-24-24a12,12,0,0,0-17,17L203,60h-2.09a76,76,0,0,0-57.2,26A12,12,0,0,0,144.87,102.89Zm-33.74,50.22a12,12,0,0,0-16.93,1.13A52,52,0,0,1,55.06,172H32a12,12,0,0,0,0,24H55.06a76,76,0,0,0,57.2-26A12,12,0,0,0,111.13,153.11Z"), search: P("M232.49,215.51,185,168a92.12,92.12,0,1,0-17,17l47.53,47.54a12,12,0,0,0,17-17ZM44,112a68,68,0,1,1,68,68A68.07,68.07,0,0,1,44,112Z") }; AG.pairSep = " vs. "; AG.reduced = function () { return !!(window.matchMedia && window.matchMedia("(prefers-reduced-motion: reduce)").matches); }; AG.h = function (tag, attrs, kids) { var svg = /^(svg|g|circle|line|text|path|rect|tspan)$/.test(tag); var el = svg ? document.createElementNS("http://www.w3.org/2000/svg", tag) : document.createElement(tag); attrs = attrs || {}; for (var k in attrs) { var v = attrs[k]; if (v == null || v === false) continue; if (k === "text") el.textContent = v; else if (k === "html") el.innerHTML = v; else if (k.slice(0, 2) === "on") el.addEventListener(k.slice(2), v); else if (k === "class" && svg) el.setAttribute("class", v); else if (k === "class") el.className = v; else el.setAttribute(k, v === true ? "" : v); } (kids || []).forEach(function (c) { if (c == null || c === false) return; el.appendChild(typeof c === "string" ? document.createTextNode(c) : c); }); return el; }; AG.money = function (v, opts) { var round = opts && opts.round; return v.toLocaleString("en-US", { style: "currency", currency: "USD", minimumFractionDigits: round ? 0 : 2, maximumFractionDigits: round ? 0 : 2 }); }; AG.pct = function (p) { var a = Math.abs(p); return (a >= 10 ? a.toFixed(0) : a.toFixed(1)) + "%"; }; AG.pctExact = function (p) { return (p 0 ? "m" : "f"; }; AG.byId = function (id) { for (var i = 0; i 0 ? "higher for the male procedure" : "higher for the female procedure"; }; AG.t = function (key, vars) { var parts = key.split("."), v = AG.copy || {}; for (var i = 0; i < parts.length && v != null; i++) v = v[parts[i]]; if (typeof v !== "string") v = key; return v.replace(/\{(\w+)\}/g, function (m, k) { return vars && k in vars ? vars[k] : m; }); }; AG.tx = function (tag, key, attrs, vars) { var el = AG.h(tag, attrs || {}); var s = AG.t(key, vars); if (/ rr.right) dx = rr.right - tr.right; if (tr.left + dx < rr.left) dx = rr.left - tr.left; tip.style.transform = "translateX(" + Math.round(dx) + "px)"; } var openedAt = 0; function show(on) { if (on && !wrap.classList.contains("is-open")) openedAt = Date.now(); wrap.classList.toggle("is-open", on); btn.setAttribute("aria-expanded", String(on)); if (on) place(); } btn.addEventListener("mouseenter", function () { show(true); }); wrap.addEventListener("mouseleave", function () { show(false); }); btn.addEventListener("focus", function () { show(true); }); btn.addEventListener("blur", function () { show(false); }); btn.addEventListener("click", function (e) { e.stopPropagation(); var open = wrap.classList.contains("is-open"); show(!open || Date.now() - openedAt < 400); }); btn.addEventListener("keydown", function (e) { if (e.key === "Escape") show(false); }); return wrap; }; AG.pairRows = function (d) { var h = AG.h; return h("div", { class: "ag-pair" }, [["m", d.mName, d.mCpt, d.m], ["f", d.fName, d.fCpt, d.f]].map(function (x) { return h("div", { class: "ag-pair-row" }, [ h("p", { class: "ag-pair-name" }, [h("span", { class: "ag-ink-" + x[0], text: x[1] }), AG.cpt(x[2])]), h("p", { class: "ag-num ag-pair-val", text: AG.money(x[3]) }) ]); })); }; document.addEventListener("click", function () { var open = document.querySelectorAll(".ag-cpt-wrap.is-open"); for (var i = 0; i < open.length; i++) { open[i].classList.remove("is-open"); open[i].firstChild.setAttribute("aria-expanded", "false"); } }); AG.stat = function (d, opts) { var h = AG.h, s = AG.side(d); var num = opts && opts.exact ? AG.pctExact(d.pct).replace("\u2212", "") : AG.pct(d.pct); return h("p", { class: "ag-stat" + (opts && opts.small ? " is-small" : "") }, [ h("span", { class: "ag-stat-num", text: num }), AG.tx("span", "shared.difference", { class: "ag-stat-cap" }), h("span", { class: "ag-sr", text: " (" + AG.t(s === "m" ? "shared.srHigherMale" : "shared.srHigherFemale") + ")" }) ]); }; AG.onVisible = function (el, fn) { if (!("IntersectionObserver" in window)) return fn(); var io = new IntersectionObserver(function (es) { if (es.some(function (e) { return e.isIntersecting; })) { io.disconnect(); fn(); } }, { threshold: 0.2 }); io.observe(el); }; AG.axisEnds = function (extraClass) { var h = AG.h; return h("div", { class: "ag-axis-cap " + (extraClass || "") }, [ h("span", { class: "ag-ink-f" }, [h("span", { class: "ag-arrow", "aria-hidden": "true", text: "\u2190" }), AG.tx("span", "shared.paysMoreFemale")]), h("span", { class: "ag-ink-m" }, [AG.tx("span", "shared.paysMoreMale"), h("span", { class: "ag-arrow", "aria-hidden": "true", text: "\u2192" })]) ]); }; AG.header = function (o) { return AG.h("div", { class: "ag-hd" }, [ AG.tx("h2", o.hed, { class: "ag-hed" }), o.dek ? AG.tx("p", o.dek, { class: "ag-dek" }) : null ]); }; AG.footer = function () { var h = AG.h; var about = h("div", { class: "ag-about", hidden: true, id: "ag-about-" + Math.random().toString(36).slice(2, 8) }, [ AG.tx("p", "shared.about1"), AG.tx("p", "shared.about2") ]); var btn = AG.tx("button", "shared.aboutButton", { class: "ag-about-btn", type: "button", "aria-expanded": "false", "aria-controls": about.id }); btn.addEventListener("click", function () { var open = about.hidden; about.hidden = !open; btn.setAttribute("aria-expanded", String(open)); }); return h("div", { class: "ag-ft" }, [ AG.tx("p", "shared.method", { class: "ag-method" }), h("div", { class: "ag-ft-row" }, [ AG.tx("p", "shared.source", { class: "ag-source" }), AG.tx("p", "shared.credit", { class: "ag-credit" }), h("div", { class: "ag-ft-end" }, [btn, AG.ta(h("span", { class: "ag-logo", role: "img" }), "aria-label", "shared.logoLabel")]) ]), about ]); }; AG.mount = function (root) { var name = root.getAttribute("data-ag"); var fn = AG.components[name]; if (!fn || !AG.data) return; root.setAttribute("data-ag-mounted", ""); if (root.parentElement) root.style.setProperty("--ag-host", getComputedStyle(root.parentElement).fontFamily); root.innerHTML = ""; fn(root); if (typeof CustomEvent === "function") root.dispatchEvent(new CustomEvent("ag:mounted", { bubbles: true })); }; AG.mountAll = function () { var run = function () { var roots = document.querySelectorAll(".ag-root[data-ag]:not([data-ag-mounted])"); for (var i = 0; i 0; }).length; var sorted = AG.data.slice().sort(function (a, b) { return a.pct - b.pct; }); var sel = null, entered = false, landed = false, pos = {}; root.appendChild(AG.header({ hed: "pattern.hed", dek: "pattern.dek" })); var findIn = AG.ta(h("input", { type: "search", "aria-label": "Find a procedure on the chart", autocomplete: "off" }), "placeholder", "pattern.findPlaceholder"); var hits = h("ul", { class: "ag-pt-hits", "aria-label": "Matching pairs" }); root.appendChild(h("div", { class: "ag-pt-find" }, [h("div", { class: "ag-find", html: AG.icons.search }, [findIn]), hits])); var wrap = h("div", { class: "ag-pt-svgwrap" }); var svg = h("svg", { role: "img", tabindex: "0", "aria-label": "Beeswarm chart. " + nM + " of 55 pairs pay more for the male procedure. Use arrow keys to step through pairs." }); wrap.appendChild(svg); root.appendChild(wrap); root.appendChild(AG.axisEnds("ag-pt-cap")); var card = h("div", { class: "ag-pt-card", "aria-live": "polite" }); root.appendChild(card); root.appendChild(h("div", { class: "ag-sr" }, [h("table", {}, [ h("caption", { text: "Table 3, facility payments, 2023" }), h("tr", {}, ["Male procedure", "Payment", "Female procedure", "Payment", "Percent difference"].map(function (t) { return h("th", { text: t }); })) ].concat(AG.data.map(function (d) { return h("tr", {}, [d.mName, AG.money(d.m), d.fName, AG.money(d.f), AG.pctExact(d.pct)].map(function (t) { return h("td", { text: t }); })); })))])); root.appendChild(AG.footer()); function drawCard() { card.innerHTML = ""; if (!sel) { card.appendChild(AG.tx("p", "pattern.emptyCard", { class: "ag-pt-empty" })); return; } var d = sel; card.appendChild(AG.pairRows(d)); card.appendChild(AG.stat(d, { small: true })); } var circles = {}; function layout() { var W = wrap.clientWidth; if (!W) return; var narrow = W < 520; var r = narrow ? 5.5 : 7, gap = 1.5, D = 2 * r + gap; var padL = r + 4, padR = r + 4; var sx = function (p) { return padL + ((p - LO) / (HI - LO)) * (W - padL - padR); }; var placed = []; AG.data.slice().sort(function (a, b) { return Math.abs(a.pct) - Math.abs(b.pct) || a.id - b.id; }).forEach(function (d) { var x = sx(d.pct), cands = [0]; placed.forEach(function (p) { var dx = Math.abs(x - p.x); if (dx < D) { var dy = Math.sqrt(D * D - dx * dx); cands.push(p.y + dy, p.y - dy); } }); cands.sort(function (a, b) { return Math.abs(a) - Math.abs(b); }); for (var i = 0; i = D * D - 0.01; }); if (ok) { placed.push({ x: x, y: y, d: d }); pos[d.id] = { x: x, y: y }; break; } } }); var maxY = Math.max.apply(null, placed.map(function (p) { return Math.abs(p.y); })); var annH = 8, axisH = 22; var cy = annH + maxY + r + 6; var H = cy + maxY + r + 6 + axisH; svg.setAttribute("width", W); svg.setAttribute("height", H); svg.setAttribute("viewBox", "0 0 " + W + " " + H); svg.style.width = W + "px"; svg.style.height = H + "px"; var old = svg.querySelector(".ag-pt-axes"); if (old) old.remove(); var axes = h("g", { class: "ag-pt-axes" }); TICKS.forEach(function (t) { var x = sx(t); axes.appendChild(h("line", { class: t === 0 ? "ag-pt-zero" : "ag-pt-grid", x1: x, x2: x, y1: annH - 4, y2: H - axisH + 2 })); axes.appendChild(h("text", { class: "ag-pt-tick", x: x, y: H - 6, "text-anchor": "middle", text: t === 0 ? "0" : (t < 0 ? "\u2212" : "") + Math.abs(t) + "%" })); }); svg.insertBefore(axes, svg.firstChild); var dots = svg.querySelector(".ag-pt-dots"); if (!dots) { dots = h("g", { class: "ag-pt-dots" }); svg.appendChild(dots); } AG.data.forEach(function (d, i) { var c = circles[d.id]; if (!c) { c = h("circle", { class: "ag-pt-dot is-" + AG.side(d), r: r, cx: 0, cy: 0, "data-id": d.id }); c.style.transform = "translate(" + sx(0) + "px," + cy + "px)"; c.style.opacity = "0"; circles[d.id] = c; dots.appendChild(c); } c.setAttribute("r", r); var place = function () { c.style.transform = "translate(" + pos[d.id].x + "px," + (cy + pos[d.id].y) + "px)"; c.style.opacity = ""; }; if (entered) { if (landed) c.style.transitionDelay = "0s"; place(); } c.__place = place; c.__at = SWEEP * (d.pct - sorted[0].pct) / (sorted[sorted.length - 1].pct - sorted[0].pct); }); svg.__cy = cy; } function enter() { entered = true; var reduced = AG.reduced(); AG.data.forEach(function (d) { var c = circles[d.id]; c.style.transitionDelay = reduced ? "0s" : Math.round(c.__at) + "ms"; c.__place(); }); setTimeout(function () { landed = true; for (var k in circles) circles[k].style.transitionDelay = "0s"; }, reduced ? 0 : SWEEP + 950); } function select(d) { sel = d; for (var k in circles) circles[k].classList.toggle("is-on", sel && +k === sel.id); svg.classList.toggle("ag-pt-dim", !!sel); if (sel) circles[sel.id].parentNode.appendChild(circles[sel.id]); drawCard(); } function nearest(evt) { var b = svg.getBoundingClientRect(), x = evt.clientX - b.left, y = evt.clientY - b.top, best = null, bd = 22 * 22; AG.data.forEach(function (d) { var p = pos[d.id]; if (!p) return; var dd = (p.x - x) * (p.x - x) + (svg.__cy + p.y - y) * (svg.__cy + p.y - y); if (dd < bd) { bd = dd; best = d; } }); return best; } svg.addEventListener("pointermove", function (e) { if (e.pointerType === "mouse") { var d = nearest(e); if (d && d !== sel) select(d); } }); svg.addEventListener("click", function (e) { var d = nearest(e); select(d || null); }); function outside(e) { if (!document.contains(root)) return document.removeEventListener("click", outside, true); if (!sel || (e.target.closest && e.target.closest(".ag-pt-svgwrap, .ag-pt-card, .ag-pt-find"))) return; select(null); } document.addEventListener("click", outside, true); svg.addEventListener("keydown", function (e) { var i = sel ? sorted.indexOf(sel) : -1; if (e.key === "ArrowRight" || e.key === "ArrowUp") { select(sorted[Math.min(sorted.length - 1, i + 1)]); e.preventDefault(); } else if (e.key === "ArrowLeft" || e.key === "ArrowDown") { select(sorted[Math.max(0, i -1; }) : []; for (var k in circles) { var hit = match.some(function (d) { return d.id === +k; }); circles[k].classList.toggle("is-hit", hit); if (hit) circles[k].parentNode.appendChild(circles[k]); } svg.classList.toggle("ag-pt-finding", !!q); if (match.length === 1) select(match[0]); hits.innerHTML = ""; if (!q) return; if (!match.length) { hits.appendChild(AG.tx("li", "shared.noMatches", { class: "ag-pt-none" })); return; } match.slice(0, 3).forEach(function (d) { var b = h("button", { type: "button", class: "ag-pt-hit" }, [ h("span", { class: "ag-ink-m", text: d.mName }), " vs. ", h("span", { class: "ag-ink-f", text: d.fName }) ]); b.addEventListener("click", function () { select(d); }); hits.appendChild(h("li", {}, [b])); }); if (match.length > 3) hits.appendChild(AG.tx("li", "pattern.moreHits", { class: "ag-pt-none" }, { n: match.length - 3 })); } findIn.addEventListener("input", find); drawCard(); layout(); if ("ResizeObserver" in window) { var lastW = 0; new ResizeObserver(function () { if (wrap.clientWidth !== lastW) { lastW = wrap.clientWidth; layout(); } }).observe(wrap); } AG.onVisible(wrap, function () { setTimeout(enter, 120); }); }; })(); window.AG.mountAll();

King, the director of reproductive bioethics at Harvard Medical School Center for Bioethics and a contributor on the study, believes that when the RVU system was created back in the 1990s, gynecological services were largely ignored, leaving the specialty undervalued from the start. While Medicare and Medicaid Services have undertaken some efforts to make RVUs more equitable across specialties over the years, that undervaluation has persisted. 

That problem may come down to how those values are determined in the first place — and who determines them. RVUs are set by a committee of doctors convened by the American Medical Association (AMA), with most of the seats allocated to members of major medical societies. The committee surveys different doctors on the time and skill required for procedures in their specialties. Based on these responses, the committee recommends values for each service to the federal government, which has historically accepted nearly 90 percent of them.

While the data quality of the underlying surveys has improved over the years, it is still not very high, said Dr. Miriam Laugesen, a tenured associate professor of health policy and management at Columbia University Mailman School of Public Health. Because the committee’s process isn’t transparent, it is often unclear what evidence its valuations rest on. This leads to distortions of all kinds, she said.

“It’s not a government process; It’s not a federal advisory committee,” Laugesen said. “It’s just a committee that was put together by the [AMA]. It’s pretty unusual in terms of how policy is made.”

In particular, King said that the surveys have a very low response rate from some specialties and are only sent to a small subset of doctors. She has never received a committee survey in the 17 years that she has practiced. 

The American College of Obstetricians and Gynecologists (ACOG) has a core seat on the committee that determines RVUs. But most ACOG members are general OB-GYNs, who split their time between obstetrics and gynecology and typically operate relatively rarely. The surgeons who perform the bulk of gynecologic operations are subspecialists – gynecologic oncologists, urogynecologists, and minimally invasive gynecologic surgeons — who complete an additional fellowship after residency. They mainly belong to their own societies and don’t have a seat at the table when RVUs are being determined. 

“From a purely cynical or economic standpoint, [ACOG] has no reason to advance billing for gynecology. It doesn’t represent a large portion of the practices of their members,” said King, adding that ACOG has been heroic in its efforts to obtain reimbursement equity for its own members. (One unintended possible result of that advocacy is that pushing for a smaller pay gap for services related to pregnancy has in turn sent the message that women’s bodies are literally worth less, and are less deserving of care when they’re not engaged in the god-given duty of reproduction.)

The result is that doctors in specialties like OB-GYN who operate infrequently are answering the surveys about complicated surgical procedures that gynecologic surgeons perform more frequently. King said that her surgical schedule aligned more closely with those of general surgeons. For example, she operates two to three times a week whereas, according to her, an average OB-GYN will likely operate once a month. 

“So they’re not the right people to be answering these surveys. They [the survey committee and ACOG] don’t target the surveys to people who operate a lot,” she said. 

Part of the problem with using surveys to determine a key foundation of medical reimbursements is that the data is self-reported by doctors. This is especially relevant when more objective avenues may be available. One of those avenues is the American College of Surgeons’ National Surgical Quality Improvement Program, which is an initiative that collects and manages key surgical data to track quality and improvement in surgical fields, including the actual time spent performing surgical procedures, making it a more definitive measure than self-reported surveys.

Dr. Shitanshu Uppal, a gynecologic oncologist and division chief for gynecology oncology services at the University of Michigan, and his colleagues analyzed 901,917 surgeries using both the National Surgical Quality Improvement Program as well as the self-reported surveys used to determine RVUs. They found that median RVUs per hour vary widely between specialties. Gynecology had the third-lowest median RVUs per hour of all the specialties studied. In effect, that means that gynecological surgeons are paid less for every hour that they work. 

The researchers also found that compared to the operating times logged for real surgeries in the Surgical Quality Improvement Program, surgeons in certain specialties tended to report longer procedure times in the self-reported surveys sent out by the RVUs committee. Specialties that reported longer operating times were more likely to be awarded higher RVUs, leading to greater hourly rates of reimbursement.

According to Uppal’s analysis, urologists — a profession most often associated with procedures on men because their work routinely concerns the prostate and the testicles — reported times that exceeded the baseline data by a median of 20 minutes. By contrast, estimates from gynecologists only exceeded the actual time by a median of five minutes, meaning that urologists tended to overestimate their time spent in surgery by four times as many minutes. Generally, urologists earn much more than general OB-GYNs, who in turn earn more than other subspecialties like gynecologic surgery. Despite having access to reliable electronic records of operating time, Uppal and his colleagues note, the AMA committee still continues to use self-reported operating times, which influences their RVU recommendations, ultimately cementing disparities in female-specific procedures. 

“The system that they have now developed over time,” King said, “now creates discrimination because of its complexity.” She added that this is how systemic discrimination is perpetuated, even if systems or rules appear seemingly neutral.

How low reimbursement rates make women’s lives worse

The US healthcare system has never been especially good at putting its money where its mouth is when it comes to women’s health. Experts have been sounding the alarm on sex-based disparities in medical billing for a long time — long enough that governing authorities like the committee responsible for RVUs and ACOG have had time to respond to the concerns. 

Over the years, the gap between male-specific procedure reimbursements and female-specific reimbursements has decreased slightly. But researchers of the Price and Prejudice study showed that between 2003 and 2023, there were no statistically significant improvements to the pay disparities. Comparing a smaller set of procedures across 1997, 2015, and 2023, they found that the gap had narrowed only minimally.

One common justification that ACOG provides for the pay gap is that obstetrics and gynecologic services see a higher-than-average share of patients on Medicaid, which pays doctors less than other insurers. King agreed this is a real factor, but it doesn’t explain why Medicare itself values procedures on women lower in the first place. And as her account of double discrimination makes clear, undervaluing gynecology ultimately affects women everywhere for the worse.

“I have patients who are waiting for surgery with me for a year at a time. I mean, that’s crazy.”

Dr. Louise King

Lower reimbursements in gynecologic surgery disproportionately affect female surgeons because they represent a majority of this specialty. It’s not just the surgeons’ salaries that are affected. Hospitals and facilities work on a revenue-based model — if gynecologic surgeries are not bringing in enough money, institutions tend to allocate fewer resources to those procedures. That, in turn, can limit access to operating rooms, surgical staff, and operating times. 

“I have patients who are waiting for surgery with me for a year at a time. I mean, that’s crazy,” King said. As a minimally invasive gynecologic surgeon, she surgically treats women with endometriosis, fibroids, cysts, or other such conditions.

It also affects the training that gynecologic surgeons receive. When there aren’t enough resources to operate regularly, opportunities to train residents are rarer than they should be. The result is that low reimbursement rates have contributed to a workforce of low-volume gynecologic surgeons who are sometimes not as well-trained as other surgeons who operate frequently.

The relationship between volume of surgeries performed and patient outcomes is well-documented in other fields. Research shows that surgeons who operate infrequently have worse outcomes, including more complications before, after, and during surgeries, compared to surgeons who operate regularly. 

Low-volume gynecologic surgeons — who perform a particular procedure less than 12 times a year — have significantly higher rates of causing bowel and urinary tract injuries in their patients. Especially in benign gynecological surgery (procedures not related to cancer), having surgeons who operate frequently also increases the number of minimally invasive procedures performed, which is related to better patient outcomes. 

Payment reform is an uphill battle, but not a lost cause

It takes multiple years and pressure from different entities to update or add new RVUs for procedures — but crucially, it is not impossible. The first few steps to creating a sustainable system might look like submitting gynecology-related procedure codes for reevaluation to the AMA committee or prompting a congressional report on the matter. 

For example, for years there has only been one billing code for endometriosis, regardless of the complexity of surgery or technique used. This disincentivizes surgeons from taking on complicated surgical cases of endometriosis because of the low reimbursement rate. Reevaluating that rate or diversifying the various surgical approaches that different cases require would help affected women get the care they deserve.

Medicare’s payment system also has what’s called a budget neutrality rule, which dictates that significantly increased payments for any given procedure must be offset by cuts elsewhere. The AMA itself believes that reforming this rule would make it easier to work toward reimbursement equity for female-specific procedures.  

Experts have suggested creating gynecology-specific conversion factors to bring parity to a system that naturally undervalues these procedures, while others think that eliminating sex-specific billing codes whenever possible is the right move. For example, the code for genitalia biopsies is different for men and women. This means that those procedures can be reimbursed differently even though they are essentially similar. Adopting one sex-neutral code for any genitalia biopsies would avoid this. While the reform required for this would be substantial, researchers believe that it would narrow the billing gap significantly.     

Authorities often justify lower RVUs for female-specific procedures by pointing to training: Gynecologists spend fewer years learning to operate than urologists and other higher-paid surgical specialists, according to King.

Historically, surgeons who operated on the female pelvis trained in ways that were more closely aligned with general surgeons than OB-GYNs. When obstetrics and gynecology merged into one specialty in the 1900s, the training for gynecologic surgery was truncated. 

“General OB-GYNs will train for four years and of that four years about 18 months is devoted to surgical training in gynecology,” said King, contrasting the new training period to those of general surgeons, which is in years. 

She believes that training periods should be increased for gynecologic surgeons. That change would secure better reimbursement for the speciality, while also producing better trained surgeons who would be more attuned to advances in minimally invasive surgery that will ultimately improve the outcomes for their female patients. In some cases, the experience gained during a two-year fellowship in gynecologic surgery was equivalent to 19 years in practice.  

In their initial work on double discrimination, Watson and King argued that failing to implement solutions for parity is tantamount to saying that female patients deserve less well-trained surgeons than male patients.  

As with most women’s health issues, it is unclear if this systemic disparity in medical reimbursement is the result of conscious misogyny or a by-product of ignorance. But that question of origins matters little when its consequences have been enshrined in medical policy, with lasting consequences for the practice of medicine more generally. We have allowed this wage gap to persist, even as many have sounded the alarm for decades. It is long past time to heed their warnings.

October 7, 2026, 5:50 pm ET: This story was originally published on October 2 and has been updated to include the sourcing on the RVU gap resulting from training for gynecologic surgery.  

Buried beneath every American city is a mysterious jumble of cables, pipes, and gas lines, entangled like a pair of wired headphones the moment you pull them out of your pocket. Every time a contractor sticks a shovel in the asphalt, they embark on a high-risk excavate-by-dots obstacle course, dodging color-coded clusters of spaghetti-like power lines, gas mains, sewers, and fiber-optic cables. A line of blue spray paint means water. Red means power. Yellow is gas. Orange is telecom. 

About 700,000 times a year — that’s once every 47 seconds — something goes wrong. 

A backhoe slices into a cable and cuts off a 911 call center, or a woman comes home to find her dryer and hot water heater on fire after a contractor pierced her electrical line. Just last week, a contractor accidentally severed a single telecommunication cable that proved critical enough to ground flights at some of the nation’s largest airports on the eve of the UN General Assembly, ensnaring even world leaders like Canadian Prime Minister Mark Carney in delays. 

The ground beneath our city streets has never been so crowded, driven by a furious race to lay fiber optic cables carrying the data that fills your social media feeds and transmits your ChatGPT queries. Nor have the systems we use to find out where it’s safe to dig — namely, a hotline and all that colorful spray paint — ever felt so catastrophically insufficient. 

“If the locator mismarked the line, if the line wasn’t marked, if there was no map that even said the line was there,” said Sarah Magruder Lyle, president of the Common Ground Alliance, a trade association focused on underground damage prevention, “there are a variety of issues that could happen before the excavator even puts a shovel or a backhoe in the ground.”

According to Common Ground, such incidents cost the US about $83.2 billion per year, roughly as much as we invest in public transit annually nationwide. But the actual financial toll may be even higher, because every time a crew stumbles across a mislabeled power line or a gas pipe everyone else forgot about, it can delay construction for months on end. As Magruder Lyle put it, “if they get out there and they find something that they’re not supposed to, you have to go back and reengineer the whole project.” 

Such surprises can also lead to tragedy. A 5-year-old named Alistair Lamb was killed in Missouri last spring after a contractor drilled into a gas line near his home, spewing fumes that leveled his house when his unsuspecting father flipped a light switch. And just this May, another unmarked gas line killed three people when contractors struck it, leading an apartment complex in Dallas to explode.

Yet the messy underbelly of our cities remains largely invisible, with accountability for issues foisted upon and at times evaded by a hodgepodge of finger-pointing private utility companies, agencies, and contractors. And that’s a shame because there is, in fact, a better way to bury the cables and pipelines that our modern lives depend on, and a multitude of better ways that we can find the ones we’ve already buried. 

What Disney World gets right about the stuff we bury

To understand one utopian vision for how all this could work, look no further than the happiest place on Earth. That’s a moniker Disney World has earned, in my mind, not for its Mickey Mouse waffles (though my niece may disagree), but for its world-class utility corridors, also known as utilidors.

Directly beneath the Magic Kingdom is a sprawling network of tunnels. Initially designed to shepherd around Mickey and his friends, the system now also serves as one of the largest utility corridors ever created. Unlike the rest of America, where cables are typically dumped in ditches before being covered with dirt, the electrical lines that power Cinderella’s castle quite literally get the princess treatment. 

Alongside water pipes, sewers, and animatronic control systems, these power lines lace the walls and ceilings of the utilidors, meaning they’re protected from weather and changes in the soil. They are also easy to monitor, catalog, and access if something needs fixing. While the cost of building a controlled system like this is much higher upfront than burying directly in the ground, the practice virtually eliminates the risk of accidentally rupturing a gas main or slicing a fiber optic cable during routine maintenance. Once they’re built, they can also drastically reduce costs in the long run because you no longer need to tear up a street every time a utility needs fixing. 

“Obviously, it’s expensive,” said Debra Laefer, a professor of civil and urban engineering at New York University, “but in the long term, it would be hugely cheaper,” especially “in areas where we are digging all the time, there’s a direct business case for moving towards these utilidors.”

As fantastical as Disney World’s imagineered system may seem — and it was, to be clear, predicated on a not-easily-reproducible master plan — the idea of storing utility lines in easy-to-access tunnels is not just for amusement parks. It is possible even in places where most of those lines are currently lying in the dirt, as Prague proved back in the 1970s, when the city decided it was fed up with destroying its precious cobblestone streets every time a buried pipe or cable needed fixing. The city now boasts a sprawling utility tunnel network, occasionally patrolled, evidently, by a couple of robot dogs tasked with mapping out every cable. 

Singapore has also spent years building out walkable utility corridors that operate much like the Magic Kingdom’s, transmitting water, data, and energy from the safety of well-organized underground tunnels. Some of Tokyo’s largest districts store their cables and gas lines in huge concrete utility ducts, and many Parisian water lines and fiber optics are kept across the city’s centuries-old sewer system, a catacombs of waste made easily navigable by blue-and-white street signs that correspond to the avenues above. Many university campuses across the US also boast their own sprawling utilidors.

In most cases, cities chose not to relocate utilities all at once, a feat that would probably be financially prohibitive and deeply disruptive. Rather, they took advantage of all the times they’ve needed to relocate utility lines anyway, a standard (and expensive) practice for big building projects like the Wallis Annenberg Wildlife Crossing or the Second Avenue Subway, a major transportation project in New York City that’s been consistently delayed by unmapped utility lines. A group of graduate students found that building such a tunnel in parts of NYC would likely cost at least twice as much as dropping a utility line in the ground, but could recoup the costs within 47 years. 

“There are streets in New York City that have been excavated at least once a year, every year for the past 30 years,” Laefer said. If you just prioritize those streets, gathering the piping in a “big concrete box” underground, you could allow workers to complete most repairs without needing to disturb the streets or sidewalks above at all.

But building new utility tunnels would require that we first come to better understand the wires and piping already entangled under our cities, which if you recall, we have only a surprisingly flimsy picture of in most cities today. But if every surprise utility line is an opportunity to move it somewhere better, it is also an opportunity, at the very least, to map out where it is.

To that end, a few years ago, Laefer led a $1.2 million pilot funded by the National Science Foundation to create a test case for 3D-mapping New York City’s underground utilities, a project that’s since been adopted into a citywide initiative. The city says it plans to roll out the spruced up utilities locating system in 2028. The UK debuted a similar digital map of over 3 million kilometers of underground pipes and power, owned by hundreds of utilities and agencies a few years ago. The nation projects that it could help save over $500 million per year by minimizing delays and accidental damage. In places where Singapore has not yet stuffed its utilities into walkable tunnels, the city’s own mapping and utilities coordination has already saved it an estimated $230 million in avoided clashes since 2024. Not bad for an intervention that doesn’t even require any physical rebuilding.

So why hasn’t every American city copied London or Singapore?

Well, US cities tend to be plagued by even more private utilities and patchwork local and state norms and regulations than a nation like the United Kingdom. Getting them to pool together what they often consider proprietary underground data is challenging, and few local officials have opted to prioritize such efforts. But this may well change as more and more fiber optics go into the ground, more flights get canceled, more apartment complexes explode, more power lines flicker off, more sinkholes form, and more Americans come to be all too acquainted with the perils lurking just beneath our feet. 

“I think at this point, it’s more about political will and money than any technical obstacle,” Laefer said. “Technically, we’re there. It’s just a matter of doing it.”

Welcome to The Midterms, Actually. Every week I’ll be writing about two things: one big idea or trend I see forming across the 2026 primaries, and one thing I’m hearing from a voice on the ground. You can sign up here to get it in your inbox every week. Let’s get into it.

One big idea: Democrats’ Laken Riley vote is aging badly

Two things happened this week that reveal how quickly the politics of immigration enforcement are shifting.

First, Republican Sen. Jim Justice of West Virginia, a staunch ally of President Donald Trump, urged the administration to dial back its enforcement push. “Before you know it, you get people that get carried away in their job and whatever it may be, and all of a sudden, then you’re dragging people away that have lived on your street corner for 15 years,” Justice told Politico this week. “I don’t know that we need to be doing that stuff.” And he’s not the only red-state Republican who’s been pushing back on ICE lately, including some facing voters in tough races next month.

The second thing flew under the radar: Sen. Jon Ossoff, the Georgia Democrat running for reelection who has also been floated as a 2028 presidential candidate, struggled to explain himself when a reporter confronted him on camera about his vote for the Laken Riley Act in January 2025. Passed within days of Trump’s inauguration and signed nine days into his second term, the law requires ICE to detain undocumented immigrants who are merely arrested for or charged with theft-related crimes, before any conviction.

In the exchange with a reporter from Drop Site News, Ossoff talked about his Senate work investigating alleged ICE abuses, but didn’t directly address his vote after being asked if he regretted it. “Laken Riley was my constituent and what happened to her should happen to no American, to no person in this world,” Ossoff said. “I think that immigration enforcement should be focused on those who are convicted or charged with serious crimes.” He excused himself to walk into a meeting as the reporter pressed him on whether the law undermined due process. 

In both cases, you see changing winds of electoral politics. Back in 2025, when Ossoff and 11 other Senate Democrats joined Republicans in supporting Trump’s immigration bill, they implicitly validated Trump’s premise on immigration enforcement: that not only was the border out of control, but that there also needed to be new efforts to step up interior enforcement through expanding agencies like ICE. In the year-plus since that vote, Trump has hired more ICE officers, launched sweeping raids in cities like Minneapolis, Chicago, and Los Angeles, and fueled a cauldron of chaos that’s led to shootings across the country. Politically, it has now put Democrats like Ossoff in a pickle, one that will matter more the closer we get to 2028.

How do you explain giving more immigration powers to a Stephen Miller-led White House? How can one argue you couldn’t have seen this coming?

In some ways, you can understand the political logic of the time. Ossoff, who is running in a state Trump carried, and many of the congressional Democrats who supported the Laken Riley law were in competitive races, and Democrats left the 2024 election thinking they had misread the electorate on the issue, particularly after Latino voters swung hard toward the right. But given what’s happened since, it’s clear many congressional Democrats made a familiar mistake: prioritizing political triangulation over standing on their values of protecting marginalized communities.

Take a look at some of the Democrats who have already had to walk back their votes. In the Democratic primary for Minnesota’s open Senate seat, Lt. Gov. Peggy Flanagan made US Rep. Angie Craig’s vote for the law a key focus of her campaign, calling it “disqualifying.” The attacks were so effective that Craig eventually expressed remorse, saying, “I wanted to be honest with the people of Minnesota. I will tell you when I get it wrong, and I got that one wrong.” Craig lost the August 11 primary to Flanagan.

Maryland’s April McClain Delaney, another congressional Democrat, also said she regretted her vote: “I will say that this was in the beginning of my term. And I never thought that ICE immigration and the horror of what has happened would happen.” (She survived her primary, defeating David Trone by 6 points.)

It’s easy to see where the buyer’s remorse is emanating from. Trump’s ICE is broadly unpopular. And as we’ve talked about on this Substack and in our reporting on America, Actually, Trump is doing worse and worse with Latino voters, with both prices and immigration enforcement driving the slide. A Hart Research poll for TelevisaUnivision of 1,500 Latino registered voters in battleground House districts (August 6–17) found 63 percent disapprove of Trump’s job performance and 62 percent disapprove of his handling of immigration enforcement and deportations. Sixty-four percent view ICE unfavorably. 

When I recently talked to Stephanie Valencia of Equis Research, she said this: 

The people who feel most regretful who voted for Trump in 2024 are most regretful because of immigration. They feel like there was a broken promise there. Like, they were told he was gonna focus on the border, which everybody agreed was chaos, and he was gonna focus on violent criminals and serious threats, and at the end of the day, he has not complied with that.

But while the shift in public opinion helps explain the elected officials’ political logic, it doesn’t change the material harm of their votes. Just this week, ICE announced plans to hire 5,000 more employees, most of them deportation officers, by the end of the year as enforcement continues to ramp up.

And these raids have a lasting effect on targeted communities, like in Maine, the Chicago area, and Texas. It’s the No. 1 issue that has come up in my conversations with local reporters. The chief architect of this chaos is the president, but make no mistake: Trump had help from shortsighted Democrats.

One thing I’m hearing 

Maya Miller is a Report for America corps member covering democracy and government for the Gulf States Newsroom, a public radio collaboration that includes Louisiana’s WWNO and WRKF. She’s been reporting on Mississippi’s redistricting fight and what residents are saying about their representation and the upcoming midterms.

You’ve been reporting from redistricting hearings in Mississippi. What are residents most concerned about, and is there a conversation or moment that captures the mood?

Residents are most concerned about their representation. In Mississippi in particular, race and party are strongly intertwined, and Black residents worry that by being drawn into new districts, voters may elect a Republican candidate. There was a bit of secrecy around redistricting earlier this summer, with Gov. Tate Reeves, a Republican, canceling a special session to redraw judicial lines. So advocates created the People’s Hearings, a series of meetings held across the state to voice concerns and create a public record of testimonies from Mississippians. Quickly after, lawmakers decided to host their own public hearings, and at these hearings, hundreds of residents showed up, mostly to urge them not to redistrict.

“It seems to me that the threat of losing representation has made people more aware of the moving parts.”

At one hearing, I spoke to Joyce Arceneaux-Mathis, president of the Natchez NAACP. She’s frustrated that the state is going to spend millions on new maps when people are struggling with day to day expenses. And Ramona Sanders of Meridian, Mississippi, has a lot of concerns on whether mid-decade redistricting is illegal. While it’s not, she says that lawmakers are opening themselves up to legal action. I haven’t heard anything from voter advocacy groups that points to future lawsuits, but only time will tell, if and when lawmakers redraw the maps, if these maps are constitutional or not.

Redistricting is often covered as a fight over which party wins more seats. What would these changes mean for the communities you cover and their ability to have a voice in Congress?

One change that could happen is that residents who are Democrats might lose the state’s only Democrat leader. Out of four congressmen, there’s only one Democrat, Bennie Thompson, who represents District 2. His district spans much of the west, from the capital city of Jackson to most of the Mississippi Delta and southwest Mississippi. It’s a heavily Black district, somewhere around 63 percent Black. He’s held this seat for 33 years and has brought millions to the district for health care, infrastructure, and agriculture. 

So residents are worried that Republicans are trying to edge him out and replace him with a more conservative leader who may not have the same concerns as the community. 

How are these fights affecting people’s interest in the midterms? Are you hearing more determination to vote, more doubt that voting matters, or something else?

People are fired up. They’re incredibly engaged. I’m hearing from a lot of people who are more determined to vote in November than when I covered redistricting in Alabama in 2024. It seems to me that the threat of losing representation has made people more aware of the moving parts. So they’re attending hearings, participating in rallies, registering to vote and checking their registration, phoning a friend and making plans to be at the polls on November 3.

What is the national conversation about redistricting missing that becomes clear when you spend time in these communities?

I think the national conversation is missing the voice of the community. It’s one thing to report on the politics of redistricting, but there’s a layer of real concern from residents that shouldn’t be ignored. They’re worried about hospitals closing. They’re worried about public schools losing funding. They’re worried about the rising cost of groceries and gas. I find that talking to and listening to residents helps me understand a community more than calling up a lawmaker and asking them their thoughts. There’s a place for that side of the conversation, of course, but I’ve always tried to center people who will be most impacted in my reporting, and I feel that leads to a better story for our listeners.

After moving to West Virginia last year, Steve Kuo signed up for a new insurance plan that would cost him $500 a month. But there was one upside: He knew he was entitled to a free preventive visit. He wanted to take advantage of it, so he made an appointment at a clinic near his new home.

But then when he got there, they asked about his medical history and he mentioned prior kidney problems. When he received a bill a few weeks later, he had been charged twice for the same appointment: Once as that free preventive visit, once as an “office sick visit” because of the kidney conversation. He was flummoxed — and very annoyed, as he shared on social media. “No copayment means no copayment, exactly as stated in the insurance guidelines,” Kuo told me over email. “What is written and agreed upon should be honored.”

He is far from alone. On Reddit, I easily found more than a dozen examples in just the past year of people getting a surprise bill for what they thought was going to be a free visit. 

One woman posted a nearly $1,000 bill she’d received for what she thought was going to be a preventive gynecology visit. Another woman said that she ended up facing roughly $3,000 bills because she was given an ultrasound (which she said she didn’t request) at a preventive visit, and then her blood draw was sent to an out-of-network lab. Another person lamented that they had been billed $170 for an office visit because they had briefly mentioned back stiffness.

It’s not supposed to be like this. Americans don’t get much when it comes to our healthcare, but most of us are supposed to be entitled to at least one thing: Once every year, no matter your health insurance, you can go to the doctor with the goal of preventing future health problems by discussing important vaccinations, cancer screenings, and cardiovascular health screenings. The Affordable Care Act made that the law in 2010.

But more than 15 years later, people all over the country are walking into their doctor’s office, thinking they’re entitled to that one free visit every year, and walking out being billed for a sick visit, blood labs, and more. Researchers estimate that Americans spend tens of millions, maybe hundreds of millions of dollars every year on services that are supposed to be free under the ACA.

Dr. Yalda Jabbarpour, a family physician in the Washington, DC, area and medical director of the Robert Graham Center for Policy Studies, told me she still deals all the time with people’s confusion about what a preventive visit really means.

“Every single time I’m in the clinic, this issue comes up with a patient. And it’s sad that it comes up,” Jabbarpour told me. “In my mind, the blame is not to be put on the patient. The blame is not to be put on the physician. Really, this points to a bigger system issue that we need to solve.”

How could a seemingly simple promise of one free annual doctor’s visit result in a situation so convoluted that patients and doctors are still struggling with it nearly two decades later? And how might we fix it?

Patients are getting surprise bills from their doctors — and they’re pissed off

The goal of the ACA was to expand coverage and make sure health insurance would be comprehensive to actually help people stay healthy. As part of that project, the authors of the law added a mandate for insurers to cover those services, including an annual wellness exam. The problem is, they did not define what the parameters of that preventive visit actually were, said Alex Hoagland, a healthcare economist at the University of Toronto. 

“It didn’t establish a clear definition of…all the diagnosis codes and procedure codes and how frequently you get that combination within a year,” he said. “They don’t define what actually is preventive care and what’s not.”

That has left each insurer to define a preventive visit on their own — and that’s where the perverse incentives that leave patients in a lurch begin. Preventive visits are supposed to be reserved strictly for screenings and talking about other preventive measures you can take to protect your long-term health, but you may get billed for an “office visit” or a “sick visit” if you talk about any specific symptom you’re experiencing, the same as if you had come in because you had the flu. 

Health insurers have an incentive to keep the definition of preventive care as narrow as possible, so they have to pay fewer claims without any contribution from the patient. At the same time, doctors want to be paid for the actual services that they are providing. If somebody does bring up a bigger health problem during a preventive visit, then the doctor reasonably wants to be paid accordingly. 

“I know, particularly in the healthcare space, it’s really popular to blame one party or another, and here I don’t think it’s a matter of corporate greed or physician incompetence or whatever,” Hoagland said. “It’s just a natural mismatch of incentives.”

So the reality is that “one free doctor’s visit” sounds straightforward, but, in practice, it’s much more complicated. You as the patient might think of it as one visit, but in our fee-for-service healthcare system, where providers bill based on each specific thing they do, it’s not. The conversation about your vaccination may be free — but asking about your back pain is not. If you get blood tests ordered, that’s another line item. The list goes on. “It’s not like you walk into a doctor’s office and it’s just one thing that happens, right?” said Sabrina Corlette, co-director of the Center on Health Insurance Reforms at Georgetown University. “It’s multiple different things that are happening when you go into a doctor’s office.”

Part of the problem may be that it’s one free visit. We Americans are consumers at heart: Tell us we’re going to get something for free and we want to take advantage of it. But having just one free visit may also lead to people bringing a laundry list of questions to that preventive visit, because they think of it as their only chance to ask everything of their doctor. 

That makes it easy for the discussion to tip over into a sick visit — and not only does that lead to a surprise bill, it can also crowd out the conversation about prevention that the visit is supposed to be focused on.

“What ends up happening a lot…is that a patient comes in for their annual visit, and in their mind, that is their time to bring up every single thing that’s been happening during the year, because they see it as the free visit to bring up everything that’s happened during the year. For some patients, that takes up the entire visit,” Jabbarpour told me. “They leave, and it’s like, ‘Oh my god, we forgot to do all the screening things: the vaccinations, the cancer screenings, and the cardiovascular risk screening.’ And then something happens down the line because you never address those screenings.”

A 2021 study by Hoagland and Paul Shafer, both of Boston University at the time, estimated that Americans with employer-sponsored insurance had in 2018 alone paid between $75 million and $219 million in medical bills that should have been free. They were more likely to be charged for things like contraception or cholesterol and diabetes screenings, with individuals sometimes paying hundreds of dollars for those services.

Hoagland told me he had been billed for preventive visits because he has type 1 diabetes and the system automatically categorized his appointment as a diabetic visit. Earlier this year, he co-authored a study that found people with chronic conditions were more likely to be charged for preventive care. Corlette said that she too had recently been charged for what was supposed to be a preventive visit after she mentioned an issue as an aside. And these are folks who study health care for a living.

“It was literally maybe a 10-second conversation where I would ask the doctor a question about something,” she said. “And that triggered it.”

Colonoscopies are another dramatic example of how this issue plays out in real life; Corlette told me the procedure was one of the first services where this fuzzy line between prevention and treatment under the ACA became obvious. People go under for what is supposed to be a preventive colonoscopy. But then their doctor finds a polyp and decides to remove it — good clinical practice, but as a result, their patient could end up being charged thousands of dollars because the colonoscopy was no longer preventive once that polyp was snipped.

“All of a sudden, that would trigger what the patient thought was a preventive service into a cost-sharing obligation,” Corlette said. “Colonoscopies can be thousands of dollars. That was when we were first like, ‘Oh, yeah, this is a more complicated area than I think the lawmakers that drafted this provision originally envisioned.’”

How the US could fix our preventive visit headaches

The ACA’s preventive visit promise was clearly well-intentioned and it’s worth remembering that, if you look in the aggregate, more Americans are getting more free preventive care than they did before the law was passed. The problem, as much as anything, is expectations: People think they’re entitled to a free visit and are understandably frustrated when they get billed for it.

“The difference is before the Affordable Care Act, maybe they anticipated some kind of cost-sharing,” Hoagland said. “But after, they expect it to be free.”

It doesn’t have to be this way. Corlette’s team at Georgetown University recently released a proposal that Americans be entitled to three primary care and two urgent care visits every year with no cost-sharing. This approach has two virtues, she told me: It will help people feel like they are actually getting value from their insurance — the very complaint that Steve Kuo raised — and, because we know primary care helps prevent problems down the road, it should also improve people’s health in the long term. Take the pressure off that one free visit and let people see the doctor a few times a year without worrying about being charged.

“In employer-based insurance, a lot of people feel like, ‘What the hell am I getting for my premium each month?’” Corlette said. “We wanted people to feel like they’re getting good value, and we wanted to get to better population health.”

Hoagland also endorsed the idea of at least one free physician’s visit every year, no matter what’s discussed. In the long term, Jabbarpour said she’d like to see copays eliminated entirely for primary care. 

But any payment reforms are not going to happen immediately. What can we do right now?

For one, doctor offices could make an effort to be as transparent as possible. Jabbarpour’s MedStar system has started calling people a couple of days before their appointment with a brief script that tells the patient that they could be charged if the conversation with their doctor touches on additional medical problems. If your doctor’s office doesn’t proactively bring it up, ask when you schedule your appointment or when they call you to confirm it. Worst-case scenario, ask your provider what they consider preventive care when your visit begins. You might also want to educate yourself on what services are generally considered preventive and should be available for free: Your insurer may make that information available on their website, or KFF has a database you can peruse.

It’s not a perfect solution. “What ends up happening is patients come in, and they say, ‘I know I’m not supposed to talk to you about anything else, so I’m not going to tell you about anything else.’ And you don’t want that to happen,” Jabbarpour said. You should ask your doctor directly: What do you consider to be preventive care? Establish boundaries at the start. That way, you can at least be aware if you are going to get charged because the conversation crossed the line to discussing an active health problem.

And if you want to see the underlying laws change, start organizing. Contact your state and federal reps and ask them what they are doing to ensure this stops happening. Things could be different: Jabbarpour said she hopes to one day work in a health system where the cost of a single primary care visit isn’t so high-stakes, where doctors and their patients can manage their health according to their particular needs, rather than worry about how to bill for one specific conversation that covers the whole range of somebody’s health — like primary care is supposed to do.

“I think some patients can be managed by seeing us once every year, once every two years, and some patients need to see us monthly,” Jabbarpour told me. “To put all patients in the same box and say you can only see your primary care physician once a year if you want it paid for, it makes zero sense.”

This story was originally published in The Highlight. To get access to member-exclusive stories like this every month, become a Vox Member today.

The Soviet Union famously churned out tens of thousands of new apartment buildings in the 1960s, ’70s, and ’80s — some assembled in as little as 12 days — together comprising tens of millions of individual apartments. Such speed is not unique to communist command economies, or even to grim concrete mega-blocks. In Sweden and Japan today, modern, high-quality new homes can be factory-built and assembled in a matter of weeks using the same basic method — prefabrication — as the Soviets.

The US, alas, evinces no such urgency to build housing. 

In some of our highest-cost cities — San Francisco, New York, Miami — it takes, somewhat astonishingly, nearly five years to erect a modestly sized apartment building, from the time an application is first submitted to building completion. That was one of the top-line findings of a remarkable new economics working paper on housing construction timelines across 60 US cities. The gaps between the fastest and slowest cities were just as striking: In Raleigh, North Carolina, a builder can complete a 20-unit apartment building in about as much time as it takes just to approve the permit for one in San Francisco, where applications languish for more than a year. In Indianapolis, a single-family home can be built in six months from proposal to completion, less than one-eighth the time it takes in San Francisco.

Such delays have costs — in delayed housing supply at a time when America is millions of homes in the hole, in lost staff time that developers ultimately bake into the cost of the new homes, and in homes that simply never get built because the process is discouraging to anyone but the best-resourced developers. The authors, Leonardo D’Amico, Evan Soltas, and Audrey Wang, found that longer permitting lengths are correlated with higher housing prices and lower rates of housing production, though they don’t claim to have shown that these are causal relationships. The research (which has not yet been through peer review and, like any working paper, may be revised before final publication) represents the first such comprehensive dataset comparing homebuilding timelines across the US. To build it, the researchers assembled 1.1 million permits covering 2.7 million housing units between 2000 and 2025 — about one in 13 new homes permitted nationally over that period, they write. 

It probably won’t come as a shock to you that housing can be built faster in Sioux Falls, South Dakota, than in Los Angeles or pretty much anywhere in California. But perhaps the most notable takeaway (to me, anyway) from the paper was how massively home permitting times have lengthened across the board over the past 25 years. Across the cities in their sample, the researchers estimate that permitting time for a standardized multifamily building rose from 0.6 years in 2000 to 1.5 years in 2023, and, for a single-family home, from 1.6 months in 2000 to 4.9 months in 2023. 

Is there really any good reason why we should take nearly a year longer to approve an apartment building than, Soltas joked in a recent interview, “the barbarians of 2006”? Shouldn’t we be getting faster and more efficient at building housing over time, not slower — especially at a time of acute housing scarcity? 

You would think. So what is going on here?

One building, very different wait times

Permitting refers to everything that happens after a developer submits an application to build a house or multifamily building, before they’re allowed to break ground. The proposal is then reviewed for compliance with a range of different sets of rules, from building safety and energy codes to plumbing to zoning and, in some places, aesthetic design. Initial applications will often be found noncompliant and then sent back to the builder for revisions, Soltas, an assistant professor of economics at Princeton, told me.  

The gulf between housing starts in California and Texas has become something of a meme in housing abundance circles. But, for all that, we have until now lacked rigorous apples-to-apples comparisons of permitting and building times in different metro areas, Soltas said. That gap in seemingly basic knowledge, what should be “step zero” facts for the national housing reform debate, became apparent to him when he was teaching an urban economics class, which became one of the impetuses for the new paper. “I realized if I needed that information to teach my students, that it might be useful for other people,” he said. 

Rather than merely pulling permits from different cities and comparing how long whatever happened to be built in each city took to approve, the researchers used the data to estimate how long an identical, standardized project would take in each place between 2015 and 2019, keeping constant its size and the density of the surrounding neighborhood. Their uniform archetypes were a 2,800-square-foot detached single-family home on a 0.15-acre lot and a 20-unit multifamily building with 1,200 square feet per apartment.

Cities with longer permitting times also tended to have longer construction periods — the longer it takes to get an application approved, the longer it then takes to do the building itself, which compounds the gap between the fastest- and slowest-building cities (it’s possible that “cities that make the permitting process difficult also put in lots of burdens that slow the construction process,” Soltas said). One advantage of the researchers’ method is that the differences across cities can’t merely be chalked up to something like the complexity of building within already-dense environments like Manhattan — the analysis already controls for that. So the enormous gaps between different places become harder to dismiss as inevitable features of development in big cities.

The study wasn’t designed to answer the “why” of it all — those will be questions for future research. “Our paper doesn’t directly identify any smoking guns,” Soltas said. Still, some of the culprits are likely obvious — California, for example, has long subjected some housing projects to extensive environmental review requirements under the California Environmental Quality Act (CEQA). Apartment buildings, especially larger ones, are more likely to have to go through CEQA review, though they’re more environmentally friendly than single-family homes (recognizing this as an obstacle to housing affordability, California has been moving to exempt many housing projects from such reviews). 

Why are we taking so much longer to permit housing?

But why, as the study finds, is permitting becoming slower across the country, not just in the most regulated cities? “We estimate that, in 2000, our standardized multi-family project had a 71-percent probability of approval within six months of applying for a permit,” the researchers write. “In 2025, that project had a 44-percent approval probability within six months.” Permitting times have come down a bit since 2023, but it’s unclear whether that just represents the reversal of pandemic-era highs, or if it reflects real progress from the pro-housing policies that have been passed in many cities and states in recent years. 

That discovery dovetails with other economics research finding that zoning codes, which were first widely adopted in the US in the early-to-mid 20th century, have continued to become more restrictive even in recent years, making it harder to build housing. 

Without knowing why exactly permitting lengths have ballooned, it’s hard to say whether those increases have occurred for good reasons. At least some of it may have to do with the aforementioned intensification of zoning and other land-use restrictions, which, I can say from having written a lot about zoning, tend to be pretty ill-conceived and hard to justify. 

The slowdown may also be connected to expanding building code requirements. I’ve written at length about that issue, too; although ever-more stringent rules might seem like a great idea — who could be opposed to more safety? — they tend to be biased against multi-family housing and likely inflate the cost of building, all without having to show that the benefits are worth the costs. Particularly the cost in housing supply that never gets built, which has the effect of keeping more Americans living in older housing stock. And do we really think a home built to 2005 standards is dangerously subpar today?  

We might ultimately decide that many of today’s building requirements — particularly ones like stronger energy efficiency standards — are worth it, but first we’d need to determine what their full costs are, which we remain far from doing. 

At a minimum, it seems unlikely that an apartment building that’s okay to build in Grand Rapids today would be unsafe for New York. So what, exactly, is all that extra permitting time buying the city? Now, at least, Mayor Zohran Mamdani can put an exact number on the delay. 

This story appeared in Today, Explained, a daily newsletter that helps you understand the most compelling news and stories of the day. Subscribe here.

Today’s newsletter is a heavy one — so please feel free to skip it if you need to. But as the uproar over an alarming rape allegation at Cornell University from 2024 keeps building, I wanted to take a minute to look at the facts of the case and the wider scourge of campus sexual assault. (We’ll get into that second topic in just a minute.)

You’ve probably read about the case on social media, but if not, let me bring you up to speed. A former Cornell undergraduate student, identified as Jane Doe in a civil lawsuit filed earlier this month, alleges that seven men pressured her into drinking and snorting large amounts of ketamine, an anesthetic that can cause dissociation, before raping her at a campus fraternity house.

At the time, she reported the incident to campus police — but her lawsuit alleges the men were never adequately investigated. Instead, the case was handled through Cornell’s administrative disciplinary system, which meted out noncriminal punishments, such as suspensions and expulsions, and permanently shuttered the fraternity house.

The case and its handling have horrified observers online and in Ithaca, New York, where Cornell is located. In response to that uproar, the Tompkins County district attorney announced on Monday that he was reopening his office’s investigation. A nine-page statement further explained why the DA’s office didn’t file criminal charges two years ago: At the time, the statement says, the woman did not allege that she was incapacitated or that the encounter was nonconsensual. 

For many online, however, that explanation only prompted more questions, such as: Was Jane Doe ever interviewed by a specialist trained to speak to survivors? And what prompted her to go to police in the first place, if she believed the encounter was consensual? The reopened investigation will take time, but should provide some answers.

Go a little deeper

Sexual assault remains appallingly common on college campuses. When I taught an undergraduate writing class in 2024, I assigned my students a personal essay and encouraged them to choose a topic they found weighty or meaningful. In a class of 11 women, two wrote about being assaulted at school. 

That proportion surprised and horrified me, though it shouldn’t have. Surveys have consistently estimated that roughly one in five undergraduate women experiences some form of sexual assault or nonconsensual contact during college. 

Awareness around the issue has grown over the past decade or so, particularly in light of the Me Too movement and after a string of high-profile cases in the mid-2010s. But progress has been more complicated.

First, it’s difficult to know if the situation is improving or worsening, because the data around campus sexual assault is really limited. Federal law has historically required that universities disclose only the incidents reported through official channels — a small minority of all cases. 

In 2022, Congress enacted a requirement directing colleges and universities to begin conducting more comprehensive surveys on campus sexual assault every two years. That would have helped measure the true scope of the problem — arguably, a prerequisite for designing and tracking potential interventions. But the Biden administration never set up that survey, and President Donald Trump has, if anything, rolled back college sexual assault protections and slow-walked Title IX investigations related to sexual harassment and violence.

Just Monday, in fact, Trump’s Department of Education formally rescinded Biden-era rules designed to better hold schools accountable for sexual assault. The department will now revert to rules put in place during Trump’s first term, which defined sexual harassment more narrowly and gave schools more latitude over how they handle allegations. 

It’s an amazing irony, really, that those rollbacks were formalized on the same day that prosecutors reopened the Cornell case. If anything, the outrage over that investigation shows that people want institutions to take sexual assault more seriously. 

One link for later

➨ Make art a little more personal. The average museumgoer looks at any one piece of art for only 27 seconds — maybe because it’s daunting to spend any more time on the task of Serious Art Appreciation. Instead of laboring to interpret a piece the “right” way, try relating it to your personal experience and other cultural objects you know well. This is also a correct and valid way to look at art! As one artist told Vox, “I don’t need the viewer of my paintings to know exactly what I meant…it’s more about an openness to a conversation.” 

Before you go…

• Did you know…that the sun’s magnetic north and south poles reverse roughly every 11 years? It generally doesn’t have much of an impact on Earth, though it can make the auroras more visible. 
• Today’s trivia: The civil rights activist Ruby Bridges became famous at age 6 for doing what? (You can find this and other brain puzzles in Vox’s daily crossword. Look for the answer in tomorrow’s edition.)
• Yesterday’s trivia: We asked you for the name of the unrequited love interest in Victor Hugo’s The Hunchback of Notre Dame. That would be Esmeralda, whose fate was way darker in the 1831 novel than in the Disney movie version. 

To no small extent, Americans have grown accustomed to being watched.

For more than a decade, CCTV cameras have recorded a hefty portion of many people’s public movements, while ubiquitous smartphones have made us all perpetually vulnerable to unsolicited documentation (if not, social media notoriety).

Until recently, however, technical limitations constrained the reach of government surveillance (at least, in the physical realm). Police may have been swimming in CCTV footage. But using such videos to track any given person’s movements was an extremely time-intensive endeavor. To do so, an officer would need to sift through hours of footage captured on many different cameras.

Yet AI is now making it possible for the government to trace ordinary citizens’ movements with unprecedented ease and precision. Or at least, this is what the security firm Flock Safety has effectively promised — and the company’s critics have loudly warned. 

Flock provides police departments with networks of surveillance cameras — primarily, automatic license plate readers (ALPRs) that photograph each vehicle that passes them. 

The firm’s main selling point is also its chief political liability: Its ALPRs log not only license plate numbers, but also generate textual descriptions of all vehicles they detect and store this information in a nationally integrated, searchable database. This makes it technically feasible for police to monitor the travels of most any driver with a few keystrokes. 

Alas, some cops cannot be trusted with that power. In August, the Washington Post reported that at least 50 law enforcement officers had been charged or accused of using license-plate readers for illegitimate purposes, often to stalk their exes or romantic interests. Even more alarmingly, some police departments appear to have used ALPR data to surveil their own critics and protesters. 

These revelations fortified an already growing movement against license plate readers. In August alone, more than 90 cities and counties canceled contracts with Flock. And this backlash has been remarkably bipartisan. Deep blue cities like Los Angeles, Austin, Texas, and Cambridge, Massachusetts, have backed out of partnerships with Flock, while Republicans in Florida have barred local police from installing ALPRs on state highways, and the Texas GOP has paused funding for Flock cameras.

The anti-ALPR movement is responding to real problems with how that technology has been deployed. And yet, according to Flock and many police departments, these cameras are also helping to nab car thieves and locate missing persons.

All this raises the questions: Do ALPRs really have significant public safety benefits? And if so, is there a way to have our license plate surveillance and civil liberties too? 

To find out, I reviewed the latest academic research, dug through the worst horror stories in the news, and reached out to criminologists about the risks and rewards of ALPRs. Although the available evidence is limited, the technology appears to have modest — but genuine — crime-fighting benefits. And while the ALPR databases have been badly misused, this is in part because the tech has gone largely unregulated. With stringent safeguards, there may be a way to make it much harder for creeps and authoritarians to monitor law-abiding citizens’ comings and goings without forfeiting a useful law-enforcement tool.

The case for ALPRs

On a late summer day in Chamblee, Georgia, Leslie Barrera was pushing her 1-year-old son Mateo in a stroller when a car pulled up next to her. A man sprung out of the vehicle, pointed a gun at Barrera, and tried to grab the baby from his carriage. But he’d chosen his victim poorly: Barrera charged him, wrested away his gun, and tried to shoot the kidnapper, only for the firearm to jam. A woman (oddly, disguised as a man) then emerged from the car and snatched the child. The baby-nabbers proceeded to pile back into their vehicle and speed away.

With the aid of surveillance cameras, police quickly identified the car as a maroon Acura SUV. Unfortunately, the abductors had swapped out its valid license plate with a paper dealer tag. Decades earlier, that might have delayed, if not thwarted, the recovery effort. But this was 2020 — and Chamblee police had access to a Flock ALPR system. By searching through its database of vehicle photos, the cops discovered a shot of the same Acura with its registered metal license plate from days earlier. The plate led them to the suspect’s name and address — and thus, within just five hours of his abduction — to Mateo.

Flock has a lot of stories like this. The company boasts that their cameras “supported more than 1 million criminal investigations” and helped locate “more than 10,000 missing persons” last year. 

“Every single day, law enforcement across this country is using FLOCK technology to solve crime and find missing people faster,” Paris Lewbel, a spokesperson for Flock, told me. 

The limited evidence suggests that Flock’s claims about its cameras’ public safety benefits are overhyped.

But Flock’s colorful anecdotes and raw tallies tell us only so much. A Flock camera can be used in an investigation without making a decisive difference in its outcome. Without ALPR photos, Chamblee police may still have been able to tie the maroon Acura to Mateo’s kidnappers. 

To gauge Flock’s actual impact on public safety, we need (at a minimum) some quasi-experiments: When jurisdictions deploy advanced ALPR systems, do they see lower crime or higher clearance rates — both compared to their own past record and to simultaneous trends in Flock-less places?

There has been remarkably little research into these questions. But a recent study from the University of South Carolina’s Ian Adams and Scott Mourtgos sheds some light into them. 

Drawing on Flock’s installation records, the criminologists identified the month that 216 law enforcement agencies activated their first camera. They found that during the first year after Flock deployment, these areas saw an 11 percent decline in motor vehicle theft, relative to similar places that remained un-Flocked. Jurisdictions that put up the ALPRs also saw the success rate of their car theft investigations rise by 16 percent. 

These findings should be taken with a few grains of salt. For one thing, police departments will be more likely to invest in an ALPR system if they’ve got a serious stolen vehicle problem. And if car thefts in an area spike, they are liable to regress back toward historically typical levels, no matter what the local police department does. 

What’s more, agencies will often install ALPRs as part of broader anti-vehicle theft efforts, which may include deploying more officers to high-crime areas, among other things. Adams and Mourtgos’s data suggests that such simultaneous actions may be responsible for a portion of their results: Although Flock adopters solved stolen car cases at a higher rate after putting up the cameras, this rise tended to begin before the ALPRs were installed.

Nevertheless, the working paper indicates that Flock cameras do make some difference. “We’re not talking huge numbers here,” Mourtgos told me, “but it’s not nothing either.”

And there is reason to think that ALPRs have greater potential to deter and solve crimes than existing results suggest, as police departments may be deploying them in suboptimal ways. 

Consider a 2025 study of Atlantic City’s ALPR rollout. On paper, the municipality was an ideal candidate for the technology: As a barrier island, Atlantic City has only eight access points, so an ALPR network could easily capture every car entering and exiting the city. 

As one would hope, the research found that car theft declined after the cameras were erected, although the authors could not rule out other explanations. Perhaps more significantly, however, they discovered that the cameras produced vastly more leads than the department was capable of logging or investigating. Between November 2022 and November 2024, the city’s ALPRs issued 12,384 stolen vehicle alerts, virtually all of which went unnoticed or ignored by department personnel; over the studied period, the ALPR system prompted only 72 vehicle stops, of which just 52 concerned stolen cars. 

“In an ideal world, someone would be monitoring the alerts in real time and dispatching an officer, who would successfully make a traffic stop,” the paper’s co-author, Rowan University criminologist John Shjarback, told me. “But yeah, that kind of rarely happened.” 

Shjarback concludes that the department lacked enough patrol officers to respond to a significant portion of alerts, while its civilian surveillance center monitors appeared to simply miss a large proportion of serious hits. “They have a lot to do in that real-time crime center,” Shjarback said. “There’s a relatively high amount of crime and gun violence and whatnot in Atlantic City. One or two retired officers on duty just weren’t enough to do the live monitoring.” 

Similar operational flaws and resource constraints may have limited the efficacy of Flock systems in at least some of the areas analyzed by Adams and Mourtgos. Indeed, Mourtgos noted that many departments in his study lacked enough ALPRs to cover most significant thoroughfares. “Even in jurisdictions that are at the higher end on cameras, you’re still covering an extremely small minority of these roadway structures,” Mourtgos said. “So, that’s probably another reason why we see only modest effects.”

Taken together, the limited evidence suggests that Flock’s claims about its cameras’ public safety benefits are overhyped; its ALPR networks do not appear to be the “single greatest change” in policing tech “since DNA,” as the company touts on its website. But, as common sense might suggest, automatically logging all license plates on major highways and intersections seems to make it at least a bit harder to get away with stealing cars (and perhaps, babies).

Mass surveillance has its downsides

Of course, the fact that ALPRs offer some benefits to public safety does not mean that they’re necessarily worthwhile. Sewing a GPS chip into every American’s forearm would surely make crimes easier to solve and less tempting to commit. But most would regard such trackers as dystopian infringements on privacy and invitations to authoritarian control. 

And Flock’s ALPRS are much the same, in the view of their staunchest critics.

The “DeFlock” movement’s fundamental argument is that the cameras give police departments more information than they can be trusted with. One danger is that wayward cops can exploit the data to stalk their exes. The more profound threat, however, is that police agencies will use ALPR systems to deter or punish lawful dissent. 

And that hazard is not entirely hypothetical. After a writer named Canyen Ashworth published an op-ed criticizing local police in Lenexa, Kansas, that town’s police began tracking his movements with its ALPRs (apparently, because Ashworth’s column led the police chief to conclude, incorrectly, that he must have been the dastardly perp behind a recent spree of…unpermitted poster-hanging incidents).

More broadly, a 2025 investigation from the Electronic Frontier Foundation (EFF) found that more than 50 federal, state, and local agencies had run hundreds of searches related to protests or activist groups through Flock’s ALPR system. It’s not clear how many of these queries were motivated by an illiberal desire to monitor dissenters. In theory, a police officer could search for, say, “red sedan at No Kings protest” because a car of that description tried to run over protesters or committed some other criminal offense in the demonstration’s vicinity. And some of the searches tallied by EFF appear to fall into that category. But many others had no discernible purpose beyond the surveillance of activists.

We need to watch the watchmen

All this is unquestionably alarming. And if widespread misuse of ALPR networks were inevitable, then their costs to civil liberties would likely outweigh their modest (and as yet, not wholly certain) benefits for public safety.

But this probably isn’t the case. Rather, governments can likely limit ALPR’s downsides — while preserving their utility to law enforcement — by more stringently regulating the technology.

“In something like half of the states, there is no regulation. Virtually none, in our view, have adequate regulation.”

Barry Friedman, founding director of NYU School of Law’s Policing Project

Until recently, legislatures had placed scant guardrails on ALPRs. As the Congressional Research Service noted in 2024, there is “no specific federal legislative framework” governing law enforcement’s use of such cameras. And most states have no comprehensive ALPR regulations. Thus, if we’re failing to prevent police departments from misusing these cameras, it may be because most state and local governments haven’t actually tried. 

“In something like half of the states, there is no regulation,” Barry Friedman, the founding director of NYU School of Law’s Policing Project, told me. “Virtually none, in our view, have adequate regulation. The police decided to do this on their own. Them and their vendors made the decision and didn’t disclose what was happening.”

Fortunately, there is a long list of ways for policymakers to constrain abuse of ALPR data. To name just a few, governments could:

• Require police departments to delete logged location records after three weeks (or less), unless they are connected to a specific investigation (thereby leaving would-be stalkers and tyrants with less surveillance data to sift through). 
• Permit ALPR searches for only an enumerated set of law enforcement purposes, such as tracking stolen vehicles or suspected violent offenders. 
• Create an audit trail for every query documenting who searched for what, when, and under what justification.
• Mandate the use of software that automatically flags suspicious queries — such as someone searching for the same license plate over and over, without attaching an actual case number — and that temporarily locks out users who appear to be misusing the database. (Flock currently offers software with similar features, but most police departments with ALPRs are not legally required to adopt such systems.)
• Expressly forbid the use of the ALPR system for tracking people on the basis of protected speech.
• Punish intentional, unauthorized searches of the database with jail time.

Virginia’s recently enacted ALPR law includes many of these provisions, while several other states have embraced a subset of these policies.

Some may still want to get de-Flocked

To be sure, for some, even heavily regulated ALPR systems will entail intolerable costs to privacy and unacceptable risks of authoritarian misuse. The American Civil Liberties Union (ACLU) supports efforts to regulate license plate readers, but only as a second-best alternative to banning the technology outright. 

“No guardrails can completely address the chilling effect of mass surveillance,” Chad Marlow, senior policy counsel at the ACLU, said. “When people know that they are being surveilled all the time, they may not intellectually dig into the, ‘Well, what exactly is happening with this data? What’s it being collected for? How long is it being stored?’ The average person isn’t going to go through all those steps.” 

Rather, Marlow argues, many will simply see cameras and fear that they’re being watched — and fearful of those doing the watching. 

“Someone who is an alcoholic and wants to go to an AA meeting — but doesn’t want the public or any of their insurance companies or the government knowing they’re an alcoholic — and knows there’s a camera outside of the church that hosts the meeting may just say, ‘You know what? It’s really important for me to keep this struggle private. And so I can’t go to that AA meeting,’” he said. 

These are legitimate concerns, although there is little hard evidence that license-plate readers have a chilling effect on protests or personal behavior. And given how much sensitive information people publicly share on social media — or willingly generate, through their internet activity — there’s some reason to doubt that many will refrain from lawful activity, merely because they spotted an ALPR by the side of the road. In fact, license plate readers have actually been in wide use for more than a decade and, until recently, few seemed to register their existence, much less, reorganize their lives around avoiding the cameras’ gaze. 

Still, reasonable people can disagree about how to weigh ALPRs’ trade-offs. My point here is merely that governments may be able to secure modest but meaningful public safety benefits from the cameras — while greatly reducing their misuse — by giving regulation a chance. Using the ALPRs of 2026 probably doesn’t require accepting the authoritarian nightmares of 1984.

This story was originally published in The Highlight. To get access to member-exclusive stories like this every month, become a Vox Member today.

Todd Waller wasn’t expecting to spend the latter half of his 30s living with roommates. But in late 2022, following a breakup with his girlfriend, he needed to find a new place to live, quickly. 

A social worker based in the Atlanta metro, Waller started googling “efficiency apartments” and stumbled onto an ad for PadSplit, a company founded in 2017 that lets property owners rent out individual furnished bedrooms in single-family homes by the week. “It said I could get approved right away and move in tomorrow,” he recalled. There was no security deposit required.

All of a sudden, Waller found himself living with six other people and able to save a lot of money. His rent was just $180 per week, or less than $800 per month. Though he earned a decent salary, he never really had anything left over after all his bills were paid, and the new financial breathing room made the tradeoff of housemates feel worth it.

After a while, Waller decided he would be willing to upgrade to a nicer PadSplit unit, one with more amenities but still with roommates. Now he pays $1,250 a month to live on one side of a 2,000-square-foot home, with access to his own private bathroom, a private entrance, and a renovated kitchen. Three other people live on the other side of the house, and he rarely interacts with them.

“My house is awesome; I pay $500 less a month than I would if I were renting [a place of my own] without roommates, and I just was able to buy a new car,” he told me. He knows one day he’ll move somewhere else and perhaps even buy his own home, but he likes that he feels he can make that decision comfortably at his own pace.

With rents outpacing wages in most of the country, legislators are increasingly looking at co-living — sometimes called house-sharing or adult dorm rooms — as an overlooked piece of America’s housing stock. Zillow counts roughly 21 million “empty nest households” — or homes where older residents living with no children have at least two extra bedrooms. Some lawmakers want to make it easier for those homeowners to rent them out, which could put workers closer to their jobs and help seniors stay in houses they can no longer afford alone. Others are eyeing a different kind of vacancy: Post-pandemic remote work has left about a fifth of American office space empty, and there’s growing interest in turning some of it into apartments. 

Many advocates for people experiencing homelessness also argue that co-living is a needed solution, something that could help people going through emergencies avoid the scary slide into shelter living, especially models where you can move in almost immediately and without a deposit. PadSplit surveyed its renters in December and found a population that mostly works (83 percent were employed) and mostly hadn’t been to college. The median income was $32,500.

That co-living needs advocates at all is a legacy of the last century. Renting a single room in a boarding house or residential hotel was once an ordinary way to live in an American city, but in the decades after World War II those rooms became shorthand for poverty and vice, and cities responded by banning new ones and paying landlords to get rid of the old. Half a century later, many people still see sharing space with strangers as a last resort, or feel embarrassed about having roommates past a certain age. And discriminatory zoning codes written in that era are still on the books in most communities, enforcing traditional nuclear family arrangements when it comes to single-family housing, by defining “family” as people related by blood, marriage, or adoption. Those rules can be wielded against home-sharing, and against specific renters neighbors want to keep out.

But these longstanding barriers are starting to finally give. Doubling up with people you aren’t related to is less unusual than it used to be. Between 2017 and 2022, the number of households sharing space with nonrelatives grew by more than 500,000. Lawmakers desperate for anything that adds more housing supply also seem less picky about where it comes from. Companies like PadSplit are putting private capital behind the model, and a wave of new laws and litigation — much of it led by conservative legal groups arguing on property-rights grounds — is making it harder for cities to tell homeowners who can and cannot sleep in their spare rooms. 

For the first time since cities banned cheap rooms, the people trying to bring them back have the momentum, and the next couple of years will go a long way toward deciding whether co-living stays a niche workaround for people like Waller or becomes something a lot more people are able to consider.

The return of SROs

At the turn of the 20th century, as American cities swelled, immigrants, day laborers, and factory workers crowded into downtowns looking for somewhere cheap and temporary to sleep. A small room could be rented by the day, the week, or however long someone needed it. Some “boarded” in private homes while others took beds in bottom-rung hotels known as flophouses. These arrangements varied widely, but became broadly known as single-room occupancies (SROs).

The turn against them came after World War II, as middle-class white families decamped for roomy new houses in the suburbs, and SROs grew more closely associated with shoddy housing for the poor and the deviant. New York and other cities started banning SRO construction in the 1950s. By the 1970s, SROs were a regular media villain, even as they still housed hundreds of thousands of people, including the mentally ill who had been turned out by the closing of state asylums. Cities started to offer landlords incentives to convert their SROs into luxury rentals. Between the mid-1970s and the 1990s, roughly a million of these affordable market-rate units disappeared from the market across the country.

The idea that perhaps cities made a mistake in banning these lower-cost housing options has been simmering in urbanist policy circles for the last 15 years, particularly as the homelessness crisis has gotten worse. The Furman Center at NYU argued in 2018 that reintroducing SROs could help address the housing shortage. The architecture critic Karrie Jacobs argued in 2021 that cities had blamed a kind of housing for problems that belonged to the people living in it. “The notion took hold that if you eliminated the housing, the marginal people would simply vanish,” she wrote in Curbed.

Still, lawmakers have dragged their feet on actually making it easier to rent these kinds of units again, and until recently the people pushing for it had little to show beyond op-eds and white papers. What’s different now is that the fight has moved out of the world of policy papers and into places where it can produce actual results, including a lawsuit that could reach the Supreme Court.

Suing over who counts as family

In early 2022, a home-sharing startup called HomeRoom was renting out two houses in Shawnee, Kansas, a suburb of Kansas City. The company rented rooms one at a time to unrelated adults who wanted cheaper housing, and at a local city council meeting that year, Shawnee’s community development director asked whether the city should support such arrangements in single-family neighborhoods at all. Two months later, the council passed an ordinance banning “co-living groups,” which it defined as four or more unrelated adults sharing a home. 

Under the ordinance, if even one adult in a house wasn’t related to the others by blood, marriage, or adoption, the whole household counted as unrelated.

While the new law forced HomeRoom to evict its tenants, it also put Val French, a Shawnee homeowner, in a bad situation. She lived with her husband, their two adult sons, and one son’s girlfriend. Because the girlfriend was related to no one, the house was now deemed in violation of the city’s occupancy rules. Her son and his girlfriend had to move out, and French couldn’t rent the empty room to anyone else.

French and HomeRoom sued in federal court in May 2023, with assistance from the libertarian Pacific Legal Foundation. They argued that their right to choose who they live with is protected under the 14th Amendment, but they ran into legal precedent from a 1974 Supreme Court case, Village of Belle Terre v. Boraas, which upheld a Long Island village’s nearly identical rule against six college students sharing a house.

The decision came down in the same years cities were clearing out their SROs, and it has given local governments cover to decide who counts as a family ever since.

A federal judge dismissed the Shawnee suit in September 2023, citing Belle Terre, and this past April the US Court of Appeals for the 10th Circuit upheld the dismissal. A few months later, the Pacific Legal Foundation (PLF) petitioned the US Supreme Court to hear the case and reconsider Belle Terre once and for all. The precedent has already frayed in places — state courts in California, Michigan, New Jersey, and New York have struck down local “unrelated occupants” limits under their own laws — but it still governs everywhere else. 

There is “a fundamental liberty right to choose household living companions — whether as an articulation of the right of intimate association or independently — that is deeply rooted in American history and tradition,” PLF’s lead attorney on the case, David Deerson, told me. 

The Supreme Court hears only about 1 percent of the cases it’s asked to take, and even if it took this one, the plaintiffs would be asking a conservative majority to recognize a new constitutional right, which that majority has spent the past few years insisting courts should almost never do. 

Deerson’s argument is aimed at that skepticism, though. When the Supreme Court overturned Roe v. Wade in 2022, it held that an unwritten right counts only if it is “deeply rooted in history and tradition.” Deerson and his allies are arguing that living with people you aren’t related to qualifies as such because, for most of American history, boarding houses and lodgers were typical staples of family life. Belle Terre is also a short opinion written by William O. Douglas, a liberal justice whose legacy has little connection to today’s conservative justices. 

The legislative front

A Supreme Court reversal would be the cleanest fix for co-living, but advocates aren’t waiting on it. State legislatures have already shown they’ll act: Iowa barred cities from enforcing relationship-based occupancy limits in 2017, Oregon and Washington followed in 2021, and in 2024, Colorado lawmakers passed a law prohibiting cities from restricting how many unrelated roommates can live under one roof absent a clear health or safety reason. 

The Pacific Legal Foundation worked on that last bill too, and has since turned it into model legislation for other states. They colloquially call it the “Golden Girls bill” since occupancy caps like these would have prevented the four unrelated women of the hit 1980s sitcom from sharing a house. 

This year Pennsylvania lawmakers introduced their own version, with supporters citing not private property rights but the potential to ease the state’s affordable housing shortage. As legislators considered the proposal, Pacific Legal Foundation published research showing that nine of the 10 most populous cities in Pennsylvania capped the number of unrelated adults who could live together. The state House passed the Golden Girls bill this summer, though the Pennsylvania Senate version has not yet received a vote. 

The objection lawmakers hear most is the same one that animated Belle Terre — that lifting occupancy caps could turn quiet, orderly, neighborhood homes into de facto frat houses. That college partying concern killed Connecticut’s proposed Golden Girls bill this year. While the bill cleared the state Senate, it never got a vote in the House, amid reported concerns from the speaker about college students moving into wealthy neighborhoods. “What if a neighbor is unhappy about four unrelated people who are 19 years old or 22 years old living in the house?” state House Speaker Matt Ritter asked in an interview with the Connecticut Mirror. 

Pennsylvania’s sponsors tried to head off those fears by amending the bill to exclude protections where someone is enrolled at a college or university. In other words, the law would have covered Blanche, Dorothy, Rose, and Sophia from Golden Girls unless one of them signed up for a class at their community college.

Nevertheless, advocates are expanding their push. Jim Manley, state policy director for the Pacific Legal Foundation, told me he expects Texas, Florida, Pennsylvania, North Carolina, Rhode Island and Utah to take up new versions of the legislation in 2027.

And occupancy caps aren’t the only co-living target. Sam Hooper, a lawyer who has been lobbying state legislatures to re-legalize SROs, and his former colleagues at the Institute for Justice are now focused on the ROOM Act, model legislation that would bar cities from restricting SROs and other co-living arrangements, including rules that make it hard to convert offices into apartments. A version was first introduced in Rhode Island earlier this year. 

For Meagan Forbes, Institute for Justice’s director of legislation, the goal is for co-living to be treated like any other form of housing. “Other local laws can regulate things like noise and fire safety, but current regulations just go way too overboard in preventing this kind of housing,” she told me. 

Hooper, who drafted the ROOM Act, says he wanted to “take a massive sledgehammer” to existing restrictions. The resistance he runs into isn’t only the frat-house worry, but older cultural fears that SROs would recreate the poor tenement conditions of the early 20th century. “I think there’s this real problem of letting the perfect be the enemy of the good in housing policy,” he told me. “People accuse us of wanting people to live in undignified housing, but they just struggle to understand that it can be a perfectly fine option for many people at different points in their life.” Hooper credits his own London co-living experience in his 20s with introducing him to some of his best friends and his wife. 

Even so, he cited a broad political coalition backing him up, from property rights champions in conservative states to Democrats and progressive YIMBYs drawn to the housing justice and homelessness angles. There are business leaders interested in workforce housing, he said, and he noted that the AARP testified for the ROOM Act in Rhode Island. “I’m hopeful we’ll be able to keep this bipartisan coalition that hasn’t succumbed to the culture wars,” Hooper added.

Not every co-living proposal is about removing rules. DC’s Senior Co-Living Program Establishment Act, also nicknamed the Golden Girls bill, would pay senior homeowners $300 a month to rent a room to another senior, plus up to $2,500 for home modifications. 

The 21st Century ROAD to Housing Act, the federal housing package that became law in July, doesn’t touch occupancy rules. But it does fund a pilot program for converting vacant commercial buildings into housing, which could make adult dorm office conversions easier.

The demand question

While momentum for co-living is real, demand is also complicated. Americans are doubling up more than they were before the pandemic — about 8.2 million households in 2024, up 4.3 percent since 2019, according to analysis shared by Zillow. Yet the share made up of unrelated adults living together actually fell slightly, by roughly 50,000 households, while homes shared with couples or families grew by nearly 400,000. In other words, when Americans need to save on rent, they still mostly turn to people they already know.

PadSplit is betting that changes. Atticus LeBlanc recently announced his company is expanding into New York, Chicago, and San Francisco, three of the most expensive rental markets in the country.

“Does that mean things have gotten more desperate, that it’s forcing this cultural acceptance?” he asked me in an interview. “I think there’s certainly an argument to be made there. But you could just as easily make the argument that, holy cow, we are really, really wasteful with space.”

Six months ago, first lady Melania Trump walked into a White House education summit next to a faceless robot powered by artificial intelligence, which she said could one day give children personalized instruction.

“Imagine a humanoid educator named ‘Plato,’” she said. “Access to the classical studies is now instantaneous…humanity’s entire corpus of information is available in the comfort of your home.” 

Weeks earlier, her guest at the State of the Union had been a 10-year-old from Alpha School, a fast-growing private school network where children learn reading and math from AI software for two hours a day, and spend the rest of their time on workshops and projects overseen by adults the school calls “guides” rather than teachers.

The Trump administration had chosen AI as the future of American schooling, and wanted everyone to know it. But they picked a bad year to ask Americans to trust the technology with their kids.

Earlier this month, top leaders at Anthropic, OpenAI, and Google DeepMind publicly urged for an AI development slowdown, warning that companies racing to build ever more powerful systems could lose control of them, threatening not only large swaths of infrastructure but even humanity itself. Trump responded by calling the panic a “hoax” and said the US would not “hinder or stifle” the industry, because “whoever wins AI wins.” This “let them cook” mantra has met nearly every AI fear this year, from job loss to data center construction.

Yet the public is not convinced, and parents perhaps least of all. Efforts to bring AI into classrooms have run into organized opposition, and much of it is coming from inside Trump’s own coalition, setting up a fight between the administration’s pro-tech allies and conservatives who think screens have already done enough damage to kids.

In the nation’s two largest school districts, New York City has recently pulled student-facing generative AI from classrooms through eighth grade and Los Angeles blocked it on all student devices. But it’s not just Democratic-run cities that are pushing back. Earlier this year Utah barred students from using generative AI on assignments without a teacher’s say-so, and required schools to notify parents when it’s used in class. Oklahoma followed suit with a law, passed without a single “no” vote, that requires districts to tell parents which AI tools they use and what data they collect, and lets families opt out. Florida went further last week, when its Board of Education adopted rules for parents to opt their children in before they use AI at school, guaranteeing non-AI alternatives, and banning “companion” chatbots outright. 

Even at the Heritage Foundation, a conservative think tank with close ties to Trump, leaders are warning about the risks of AI in the classroom. In a March report, Heritage officials cautioned that AI tools built with addictive design features can harm children’s attention, mental health, and learning, and said schools should set limits rather than assume the technology will regulate itself. In July, the Manhattan Institute, another conservative think tank, recommended New York State bar students from using AI until third grade and keep it out of tests at every grade.

All the while, the Trump administration continues to push in the opposite direction, casting the wider AI race as one the US cannot afford to lose to China. Last year, Trump signed an executive order creating a White House task force to bring AI into K-12 classrooms, train teachers to use it, and build an “AI-ready workforce” to keep America as the global leader.

Polling data released in August found voters across the political spectrum concerned about AI in classrooms, with 85 percent worried about students using it to finish assignments without learning anything. Republicans were, if anything, slightly more eager than Democrats for government guardrails, 80 percent to 76 percent. Put differently, Republican voters appear closer to the Heritage Foundation than to the White House, which points to a coming clash inside the GOP over parental rights, innovation, and whether schools that take public money should have to prove their tactics work.

Up to this point much of the AI backlash has focused on regulating the tech in public schools, while some of its most ambitious boosters have concentrated on building private schools and developing new tools. A state can cap screen time in its elementary classrooms and let another family down the road pay thousands of dollars for Alpha, and everyone can call it a win for parent choice.

That peace may start to fray though as more public money comes on the table. Texas recently expanded vouchers to Alpha network schools and the company has also sought to open publicly funded charter schools. Other states are likely to confront these issues as well, meaning the AI education fight on the right may soon be unavoidable — and would land before an electorate whose patience for the tech, in schools and increasingly everywhere else, has been running thin all year. 

The Trump administration’s AI bet

The AI-in-schools executive order came just three months into Trump’s second term, and called for AI education to begin as early as kindergarten. Within five months, Education Secretary Linda McMahon was touring Alpha School’s Austin campus, describing it as “the most exciting thing” she’d seen in education in a long time. 

Alpha is hardly the first education company to promise a shiny new school model, but it has quickly become an influential player in the growing national debate around AI in schools. Its story so far reflects many of the big questions around AI in the classroom, including how to determine if the tech is helping or hurting kids, who is profiting, and what role governments should play in facilitating its adoption.

Alpha had been a small operation for most of its existence. MacKenzie Price, a Stanford graduate, co-founded it in 2014 after her second-grade daughter complained that school was boring. The tech billionaire Joe Liemandt became Price’s partner and the school’s principal financial backer. As recently as 2024 it served about 300 students across three campuses. 

Growth exploded in summer 2025, when Alpha took over campuses from a shuttered Montessori chain and announced plans for 10 new schools. This August it announced plans for 27 more.

Behind the school’s growth is its claim that students enrolled can cover a full day’s worth of academics in just two hours with AI, learning twice as fast as they would at a traditional school. Alpha can “enable the increase in human intelligence,” Price declared on a panel in Washington, DC, earlier this month, because “our academics can be delivered so much more efficiently and effectively.” 

Alpha says results from MAP tests, a widely used measure of academic growth, show its students learning 2.6 times faster on average than similar students at other schools. But no outside researcher has audited or independently verified those results.

That may soon change. MIT’s Blueprint Labs, the group behind some of the most rigorous charter school research in the country, told Vox it is designing pilot studies to measure the effects of the school’s programs on achievement and engagement, using experimental designs. Alpha is paying for the research; the lab says its own researchers will set the questions and methods. Harvard’s Human Flourishing Program is separately planning to study students’ life skills and well-being via a longitudinal study during the next academic year, Christina Hinton, who is involved with the research, told me. For now, Alpha’s academic claims remain its own.

Meanwhile, scrutiny of Alpha has grown. A Wired investigation last year described the school’s surveillance system as capable of recording students’ screens, keystrokes, and mouse activity and potentially tracking their eye movements. A 404 Media report published this winter claimed Alpha’s own internal documents said its AI-generated lessons sometimes did “more harm than good.” Alpha has disputed both reports, saying they relied on cherry-picked examples and accounts from fired employees, and provided me with an open letter it says every current family at the Brownsville campus signed, stating that the school Wired described doesn’t resemble theirs.

But education officials are still wary. This month, ProPublica and the Texas Tribune reported that at least nine states, Texas among them, had turned down Alpha’s bids to open publicly funded charter schools, with officials repeatedly pointing out that the model had only been tested on affluent students. (Tuition varies based on location, ranging from $10,000 a year at its Brownsville campus to $75,000 at its most expensive California schools.) Where Alpha had worked with lower-income students, ProPublica and the Texas Tribune found, it either hadn’t released enough data to back its claims of student growth or state test scores fell well short of what its leaders had predicted. 

After the Texas State Board of Education rejected Alpha’s charter bid in the summer of 2025, Alpha didn’t reapply. While policy advocacy was “frustratingly slow,” Price wrote, Alpha’s goal was “for students everywhere at every income level to be able to use this model of learning.” They turned instead to Texas’s new voucher program, which lets families spend public money at private schools. Unlike a charter, a private school in the program doesn’t have to give state tests or show anyone how its students perform. The comptroller, which runs the Texas Education Freedom Accounts (TEFA), approved most of Alpha’s Texas schools, which, counting virtual programs and specialty academies, now number more than 30.

None of the concerns that Texas’s State Board of Education had raised about Alpha’s test scores, missing data, or AI applied. “That wasn’t relevant,” Travis Pillow, a spokesperson for the TEFA voucher program, told ProPublica. “Our role is to make sure that a school meets the requirements in the law.” 

The five most expensive Alpha schools in Texas opted to leave the voucher program earlier this month, but many Alpha network schools are still participating, including Texas Sports Academy, Montessorium, NextGen Academy, Nova Academy, and Waypoint Academy. TEFA confirmed to me that students are attending Alpha affiliates via vouchers this year. 

Anna Davlantes, a spokesperson for Alpha, told me that in lieu of vouchers for their most expensive campuses, her network is offering families financial aid. 

The conservative revolt

Despite Alpha’s enthusiastic embrace in the White House, the introduction of AI in schools is increasingly challenged by forces on the right, from grassroots MAGA parents to social conservatives and technocrats.

The new AI rules in Florida, Oklahoma, and Utah, for example, grew out of a broader conservative push against screens in classrooms that predated the rise of chatbots and was turbocharged by parents’ frustrations with virtual learning during the pandemic.

Earlier this year Iowa capped elementary school screen time at an hour a day in a bill Gov. Kim Reynolds signed with Health and Human Services Secretary Robert F. Kennedy Jr. at her side. Tennessee passed a weakened version of a bill that would have banned devices in elementary schools outright, while similar bills in Missouri and Kansas died after school districts and teachers unions objected.

Uniting these efforts is a growing conviction that the last 15 years of education technology has failed most kids. The sponsor of Kansas’s bill used to sit on his state’s board of education and remembers thinking that ed tech held such promise. “What we found out is that not only is it not the education of the future, but it is destroying the education of the past,” he told Education Week. A growing number of lawmakers and parents have been citing The Digital Delusion, a book self-published in 2025 by neuroscientist Jared Cooney Horvath, that blames the proliferation of screens for declining student achievement. In January Sen. Ted Cruz (R-TX), who is otherwise one of the Senate’s leading opponents of AI regulation, invited Horvath to testify about the harms of AI on student learning and mental health.

The screen skeptics overlap with a more traditionalist wing of the GOP that objects to classroom technology on principle. Missouri’s original bill would have required cursive and kept 70 percent of student work on paper. The classical education movement, which holds up teachers as a moral example, has grown fast on the right. (Of the 895 identified classical schools nationwide, roughly one-third opened or converted to the model between 2020 and 2024, according to a Heritage Foundation tracker.)

The strand of conservatism that cuts closest to Alpha School comes from a third group, the school choice advocates. On paper, Alpha sounds like what the school reformers have wanted for years: It can offer an alternative to underperforming local public schools, its AI programs can theoretically compensate for inconsistent teacher quality, it provides personalized instruction, and it checks mastery on every lesson, so families can easily know whether their child is learning. Those are problems the education reform movement was launched to solve, and Alpha claims to fix them all.

The reformers, though, want proof. Mike Petrilli, who runs the conservative Thomas B. Fordham Institute think tank and has spent decades advocating for charter schools, wrote recently on his Substack that parents who want to pay for Alpha out-of-pocket should be able to, but “that doesn’t mean the rest of us shouldn’t be skeptical, or that we shouldn’t worry about fraud, or that we should allow public funding to flow to these school.”

This split was on display in DC this month at the education conference where Price appeared. Molly Hart, who runs Utah’s public schools and calls herself a school choice supporter, argued that once a school takes taxpayer money, “the game changes,” and parents can’t make a real choice without data about how it performs. Erika Donalds, who chairs the education center at the America First Policy Institute and is one of the Trump administration’s closest allies on schools, said parents, not the government, should decide which results matter. Donalds compared the choice to picking a restaurant on Yelp. 

The evidence has started to catch up with parental concern. In a randomized trial involving nearly 1,000 students at a Turkish high school, those given access to a GPT-based math tutor that readily supplied answers scored 17 percent lower on a subsequent exam taken without it. A version designed to withhold answers eliminated that penalty but produced no statistically significant learning gain. A newer working paper tracking nearly 27,000 Chinese students found that adopting AI raised homework scores by 18 percent but lowered monthly exam scores by 20 percent within six months.

Alpha School is working to add more detail to that pile, and other studies will surely follow. It’s at least possible that some aspects of AI in education prove more promising with further study. But public skepticism will remain a major barrier, either way. On an issue where parties are still sorting themselves out, the White House and its allies are swimming against a growing anti-tech tide that extends far beyond just AI or even schools. That’s unlikely to change before 2028, and whoever wants the Republican nomination will have to say which parents they’re listening to.

The AI freakout of the past few weeks has centered on nothing less than the risk of human extinction — on the possibility that artificial intelligence could wipe us all out. But the reality is, AI programs are already having a direct and sinister effect on people’s physical health: UnitedHealth Group, the largest health insurer in the US, has been accused of using AI programs to deny payments to people, despite their doctors saying the care was necessary, and is facing a class-action lawsuit. 

Humana and Cigna are facing their own litigation over similar allegations. Meanwhile, Medicare has launched its own AI tool that will help the government determine whether to approve prior authorizations, and doctors are sending the federal government reports of patients at their offices who are in tears because they can’t get the medicine to manage their chronic pain.   

The claims in the legal complaints are harrowing: Families are desperate for an elderly loved one who recently experienced a serious fall to be allowed to stay in a nursing home. Doctors are flummoxed when a patient who may need months of rehab is cut off by an algorithm after a couple of weeks. 

This is AI’s threat to people’s health right now. 

While the current conversation may be fixated on something more Hollywood, private insurers and the Trump administration are already deploying AI programs to vet people’s health insurance claims — and the evidence is growing that these programs are not always doing a good job and are instead putting people’s health at risk. Until these lawsuits are settled or policymakers can figure out a way to regulate how insurers use this new technology, it will be up to each of us to make sure that an AI program isn’t being used to cheat us out of insurance benefits that we’re entitled to.

“The type of care that clinicians deliver is very sensitive to the patient’s unique circumstances. That may not really show up as a number. There’s not a numeric value that tells you the completeness and connection of a patient’s social network,” said Mika Hamer, a health services researcher at the University of Maryland. “There really are a lot of factors that really can’t be quantified and so aren’t really suitable to go into these models. If we start to rely exclusively on the models, I think we run the risk of losing that human touch.”

Patients say private health insurers are using AI to deny claims

Insurers and patients fighting over medical claims is nothing new: Policyholders want the most coverage they can get and payers have a business incentive to pay as little as they can. Health insurers have also used algorithms for decades: Plug in somebody’s age, their weight, whether or not they smoke, and one of the old-fashioned algorithms could spit out a rough estimate of how much medical care they would need before they died. Simple — and comparatively transparent. 

When you submitted a claim or sought prior authorization, there was supposed to be a human doctor who could review all of the relevant information on behalf of the insurer and make the final decision, in theory with the patient’s best interests in mind. It didn’t always work that way in practice, of course, but that was the idea.

What’s different right now, according to Daniel Schwarcz, a law professor at the University of Minnesota who has looked at health insurers’ AI use, is that these newer programs can make very specific predictions about someone’s future health based on the information that’s fed into them, and those predictions are being used in the process of approving or denying medical claims. And exactly what went into making those predictions — and whether it is the AI or the human who ultimately holds the final authority — is often opaque from the outside.

“The reason we want human review is we want human substantive judgment. We want AI to be supporting that,” Schwarcz said. “But the difficulty is distinguishing between AI supporting meaningful human processing and AI basically replacing the human with the human just passively defaulting to the AI judgment. It’s just incredibly hard to distinguish between those two in practice, particularly in rules or regulations.”

And those questions are at the heart of the lawsuits brought against three major insurers in the past few years about their use of AI in processing claims.

In May 2022, Gene Lokken, then 91 years old, fell in his home. He was taken to the nearby hospital in Tomahawk, Wisconsin, and treated for a broken leg and a broken ankle. Less than a week later, he was placed in the health system’s hospice care because, according to the lawsuit filed by his family, he was “very weak, not communicative, and in constant pain.”

But a month later, after receiving skilled nursing care, Lokken started to improve. By late June, his doctor had removed his splint and placed his leg in a removable boot. In July, he started physical therapy to get up and walk again. And for a while, his Medicare Advantage plan sold by UnitedHealth covered all of the services he needed.

Until July 20. Around that time, according to the lawsuit, UnitedHealth denied any more services at the nursing facility where Lokken had been for the past two months. It insisted that he could be safely discharged and more time at the nursing care facility “was not medically necessary.”

The insurer’s decision left Lokken and his physician “dumbfounded.” The physical therapist notes from the time said Lokken was still suffering paralysis and weakness after only a couple weeks of rehab. His family appealed, but UnitedHealth continued to deny the claims, asserting that because Lokken could feed himself and did not require help to clean himself, he did not need nursing care and could go home. 

From July 2022 to July 2023, Lokken remained at the nursing home, but his family had to pay between $12,000 and $14,000 monthly to keep him there. And on July 17, 2023, Lokken died — having never left the facility.

The lawsuit alleges that the decision to stop covering Lokken’s nursing-home stay was largely made by an AI algorithm, called nH Predict, which UnitedHealth follows so unquestioningly that employees were allegedly at risk of being disciplined or fired if their claims decisions went against the AI’s projections. The complaint alleges that the AI model would typically start denying claims after two weeks of nursing home care — and the attorneys have more examples than Lokken.

Glennette Kell required emergency surgery after a fall in late August 2023 and was referred to a nursing home rehab center. By mid-September, UnitedHealth was denying her claims for nursing care, despite the fact that, according to the lawsuit, she still could not walk. Kell had to pay $10,000 of her own money to cover the rest of her bills. Darlene Buckner suffered a heart attack in October 2023 and underwent triple-bypass surgery. She was readmitted to the hospital in late November because she was experiencing complications, but less than two weeks later, UnitedHealth allegedly denied any further treatment at the hospital because it was deemed medically unnecessary. At the time, according to the class-action lawsuit, Buckner was experiencing “an open infected wound near her heart from the open-heart surgery, she was on blood thinners, and was receiving antibiotics through an IV.” 

UnitedHealth, in its responses to the plaintiffs, has asserted that federal laws and regulations already require all claims decisions ultimately be made by a human doctor, not an AI program, and raised several other technical legal objections (such as saying the plaintiffs have not fully exhausted their administrative appeal options). The plaintiffs, however, allege that UnitedHealth employees “are instructed not to deviate” from the AI model’s projections, and then tell patients the AI-generated information is proprietary when they or their doctor try to inquire about why their claims are being denied. 

And, to be clear, UnitedHealth is not the only insurer being sued for using AI programs to deny claims. Humana is facing its own lawsuit brought by other Medicare Advantage patients who allege the same nH Predict program also led to their claims being wrongfully denied. Joanne Barrows, then 86, fell in November 2021 and went to a rehab nursing facility after her hospital stay. And just two weeks later, Humana began denying her claims for further care, according to the complaint, leaving her and her doctor “bewildered” because she was under doctor’s orders not to put any weight on her leg.

And the third lawsuit, filed against Cigna, alleges the company used a different AI algorithm that allowed its doctors to deny claims in huge batches — hundreds or thousands at a time — if they did not meet certain predetermined criteria. One patient, Suzanne Kisting-Leung, had claims denied for two ultrasounds done by her doctor to check for ovarian cancer, putting her on the hook for more than $700 in medical bills because, the lawsuit alleges, Cigna’s AI program determined that the scans were not medically necessary. The plaintiffs say that their claims were instantly rejected, without their medical files ever being opened. As for the law that human doctors have to make the final call? According to the lawsuit, the algorithm “enable[s] its doctors to automatically deny payments in batches of hundreds or thousands at a time for treatments that do not match certain pre-set criteria, thereby evading the legally required individual physician review process.”

These people, many of them elderly, had no idea AI was being used to make decisions about their coverage — until it was too late.

“From their perspective, you might as well have been speaking Esperanto. They just had no idea what it was that was being said to them when they were being told. They usually heard about it for the first time when they were being told that they needed to leave,” Glenn Danas, one of the attorneys in the class-action litigation against United, Humana, and Cigna, told me. “They just had no idea what was going on or why this was being done to them.”

AI is now making decisions for traditional Medicare claims too

It’s not just private insurers either. The Trump administration is now pushing traditional Medicare and Medicaid to use AI to adjudicate insurance claims too. In June 2025, nearly two years after those lawsuits had been filed against Cigna, Humana, and UnitedHealth, the US Department of Health and Human Services announced it would deploy a new program called the WISeR (Wasteful and Inappropriate Service Reduction) model to make prior authorization decisions for about a dozen procedures for the people on traditional Medicare in six states — “leveraging enhanced technologies, such as Artificial Intelligence (AI) and Machine Learning (ML), along with human clinical review to ensure timely and appropriate Medicare payment for select items and services.” It would be a pilot program, before federal officials decide whether to take the program nationwide.

It was a quiet, massive change for traditional Medicare, which covers about half of the 66 million Americans on the program. (Medicare Advantage plans, in which private companies administer Medicare benefits under slightly different parameters and which were the subject of the earlier AI-related class-action litigation, cover the rest.) People on traditional Medicare had historically not been required to get prior authorization for most services; now they would be, and a novel AI program would play a significant role in which claims get approved and which ones don’t.

From the start, physician and patient advocates warned about the potentially adverse consequences of the WISeR program; the Center for Medicare Advocacy said the program would likely “complicate, delay and even outright deny necessary care.” According to media reports and internal government documents, that’s exactly what has happened since the model went live in January. 

The Washington Post first reported in March that physicians and their patients were experiencing approval delays or denials for medical services that should have been covered. One glitch led to a denial of payment for services that had been approved. Doctors noticed that certain procedures seemed to give the AI model problems — such as two competing treatments for incontinence — and were weighing whether or not to stop offering one of those services altogether. The Post reported on one Arizona patient who had been receiving steroid injections for chronic back pain before being denied multiple times after the AI program’s launch. The pain became so severe, he said, that he could not exercise at all.

Then this month, the Electronic Frontier Foundation released a trove of documents it received in response to a Freedom of Information Act lawsuit related to the WISeR model. The documents revealed aggressive denials — one company using the AI model had actually denied more prior authorization requests than it had approved — and there were delays even when the care was ultimately approved. One review took 83 days when the model was supposed to make a decision within 72 hours.

The direct feedback from doctors in the federal documents obtained by EFF was devastating.

“We have patients calling our offices crying in pain because their procedures are being delayed while awaiting approvals or guidance tied to this model,” one doctor wrote. “A 3-4 day delay for necessary pain procedures is already difficult for vulnerable patients, but when providers cannot obtain answers for weeks, the situation becomes unacceptable.”

According to the federal documents, the Trump administration was already discussing ways to expand the program. Despite the outcry from doctors and patients, and growing concern among their representatives in Congress, it doesn’t look like WISeR is going away.

How can we make sure AI doesn’t take over insurance claims?

AI is becoming a part of healthcare — it is already happening. And many of those uses could be truly beneficial. Anthropic recently announced an initiative to apply the theoretical findings of its AI models to real-life lab work that could lead to new drugs being developed in the future. In terms of scientific discovery and the potential to translate it into new treatments, the excitement among clinicians about its potential is palpable.

And nobody denies that health insurance administration is chaotic and burdensome and would benefit from being made more efficient and streamlined — tasks which AI is at least superficially well-suited for.

But as the past few years have already shown, rushed and sloppy deployment of AI to take responsibility for approving or denying payment for people’s medical care has created serious problems. The issue is that this technology is being adopted more quickly than any regulator can oversee it. We don’t really know how health insurers are using AI tools. We don’t have a good framework for what keeping a “human in the loop” of insurance decisions — which is what people want — would really mean, Schwarcz said. 

“I think everyone agrees at a high level of generality, ‘Yes, AI should be used to enhance human review and not to replace it,’ but no one actually knows how to write rules to enforce it,” he told me. “I don’t think that there’s anything I’ve seen that has effectively defined what it means for a human to be meaningfully in the loop in the sense of retaining independent judgment as opposed to just reviewing and deferring. It is incredibly hard to specify in rules or regulations a standard that is verifiable and enforceable. That is still, as far as I know, an unsolved problem.”

Some states have started adopting rules requiring transparency for AI use in insurance claims, and lawmakers in Congress have put forward proposals requiring human involvement, including the bipartisan Protecting Patients from Automated Denials Act. But these technologies are so ubiquitous and so flexible that rigorously regulating their use is going to be difficult. Meanwhile, the profit incentive for companies to adopt these tools and deploy them widely is extremely strong. We don’t even know how widespread their use really is.

Danas said he and his clients hoped their litigation would help with “putting the brakes” on AI’s use in insurance. Until we figure out what to do from a policy perspective, what can each of us do in the meantime? Hamer told me that the old advice is still the best advice: If your claim is denied, appeal it. An enormous number of denied claims end up getting overturned on appeal. 

“You do not necessarily have to accept that the first answer from your insurance company is the final determination,” she said.

Welcome to The Midterms, Actually. Every week I’ll be writing about two things: one big idea or trend I see forming across the 2026 primaries, and one thing I’m hearing from a voice on the ground. Let’s get into it. You can sign up here to get it in your inbox every week. Let’s get into it.

One big idea: Latinos didn’t come home. They’re just in motion.

In the last presidential election, President Donald Trump’s strong support with Latino voters ended up one of the biggest stories of the night. After years of Democrats telling themselves that the growing number of Hispanic Americans would help lead to a permanent liberal majority, their dreams were first dashed in 2016 (when many of those voters stayed home) and then buried in 2024 (when they, and other nonwhite voters, swung toward Trump).

But just as Democrats have at times been too confident that minority voters would fall in line, it now seems Republicans have replicated their mistake. Poll after poll shows a historic reversal among Latino voters in Trump’s second term, and these days even Republican members of Congress are beginning to call Trump out on his deportation efforts. 

Why? It comes back to the numbers.

As Nate Cohn observed in the New York Times this week, Democrats now lead Republicans among Hispanic likely voters by 35 percentage points on the generic congressional ballot, 64-29 — a 28-point swing from 2024, when Kamala Harris carried Hispanic voters by just 7 points. The shift is a big reason the polling has Democrats in “blue wave” territory this November.

Last week on America, Actually I explored this shift in detail — by talking to Stephanie Valencia, co-founder and president of Equis, the research firm that has spent a decade digging into the Latino electorate. And the first thing she wanted us all to know: Get rid of the idea that Latinos will ever be permanently Democratic or permanently Republican. Part of the electoral power is that a larger section of the electorate defines itself more independently.

“Latinos often get given the title, the sleeping giant, which I just want to rid that out of everybody’s vocabulary. We should be called the swinging giant based on how much shift we have seen election to election. A third of Latinos consider themselves kind of independent or apolitical nonpartisan, and that’s that swinging middle that really is deciding every single election.”

Here are three more trends Valencia pointed out, which help explain why this all-important voting group has done such a drastic 180 on Trump in his second term.

1. It’s the economy (again)

As with all things in Trump’s second term, it starts with the fundamental problem of his second term: He has not delivered on lower prices. Equis found that Trump’s net approval with Latinos on cost of living sits at -52 — and his number on the economy (-45) is worse than his overall approval with the group (-31). Both are dramatically worse than where he sits with non-Latinos, who put him at -39 on cost of living and -25 on the economy. Among Latino men 18 to 39, cost of living hits a stunning -63.

Here’s what Valencia said:

“Latinos make up more of the wage workforce than any other racial subgroup. And so what that means is when wages are down and inflation is high, Latinos feel it more. We have less savings. We have less rates of home ownership. So the kind of safety nets that other people might have, we just don’t have.”

She also pointed this out:

“In Trump 1.0, his approval on the economy and the cost of living was actually better than his everyday approval… And he has been underwater with Latinos almost this entire term so far on the cost of living and the economy. It’s almost been a complete and total reversal.”

One of the arguments Trump made to Hispanic communities (and other groups also) was that he’d economically empower them, and that this was more important than prioritizing social causes that didn’t have a material benefit. That’s also been his fundamental flaw. From tariffs to rising gas prices, he has not only not delivered — people also feel worse off than before.

2. The immigration backlash

Valencia’s data was undeniable: More and more Latinos turned on Trump as he broadened his deportation efforts, which were broadly unpopular. But what I didn’t understand is why it took until Trump’s second term for people to realize those threats were legitimate, considering they were so front and center during his campaign. (They held up “Mass Deportation Now!” signs at the Republican National Convention!)

Here’s what Valencia said:

“I think this is really important for us as avid news consumers to understand what it is that the average voter, and especially that voter who’s coming in and out of elections, the kind of news they’re consuming. What they were seeing on mainstream media, perhaps even Fox News at the time, was a focus on Trump surrogates who said he’s gonna focus on serious criminals, securing the border, and violent criminals. And so that’s what people bought.”

“Fast-forward to 2026, the people who feel most regretful who voted for Trump in 2024 are most regretful because of immigration. They feel like there was a broken promise there. Like, they were told he was gonna focus on the border, which everybody agreed was chaos, and he was gonna focus on violent criminals and serious threats, and at the end of the day, he has not complied with that.”

There’s also an element of disbelief. When Trump said he would deport “bad people,” a lot of people thought he was only talking about people in other communities — or immigrants with criminal legal histories — not the more far-reaching immigration sweeps we’ve seen since he’s been in office. A striking ad from Rep. María Elvira Salazar (R-FL) this week where she directly criticizes Trump’s immigration policy explicitly made this distinction. 

The Equis numbers track that shift almost perfectly. In December 2024, 65 percent of Latino Trump voters believed he would target only criminals and recent arrivals. Today, 62 percent of Latinos overall say he’s going after undocumented immigrants broadly — and among his own Latino voters, the share who say that has jumped from 26 percent to 37 percent. That perception is the whole ballgame: Latino Trump voters who still believe he’s targeting only criminals give him +84 net approval. Among those who’ve concluded he’s targeting everyone, it collapses to +31.

The backlash isn’t about the border. It’s about what’s happening in the interior — the raids, the family separations, the citizens and legal residents swept up. 

Equis finds 68 percent of Latinos say Republicans promised to deport criminals and deported hardworking people instead — and 67 percent say Democrats promise immigration reform and never deliver. Forty-two percent say both parties broke their promises.

“I think it is a huge mistake for Democrats to assume or to rest on their laurels and think that because of what people are feeling around Trump and Republicans and the sense of disappointment and broken promises, that that is automatically going to translate to them as a party and as a brand. Right now, the Democratic brand has suffered a repeated set of blows among Latino voters. 20 years of comprehensive immigration reform that hasn’t been delivered. It’s a thing that comes up in almost every focus group.”

3. Meeting voters where they are

The third trend isn’t about persuasion at all. It’s about the fact that a lot of campaigns are still bad at the mechanics of reaching this electorate — and Valencia’s examples were almost comically basic.

Their testing found that more than 70 percent of Latinos who see an ad in Spanish come away thinking that candidate or party cares more about Latinos. That’s not a persuasion effect. That’s a respect effect, and it’s enormous for the price of one ad buy.

So how do campaigns blow it? Some examples, according to Valencia: translating English ads directly into Spanish, rather than tailoring them to the language and culture; and assuming English speakers and Spanish speakers are separate categories when voters often shift between media in both languages. 

The deeper error is the monolith problem — the group is still treated as a voting bloc that moves in unison, rather than a diverse set of people from varying countries of origin and contexts. And the fastest-moving division isn’t the country of origin at all. It’s age — an electorate replenishing itself faster than either party’s model of it.

“There are a million Latinos who turn 18 every single year, and an 18-year-old who turns 18 this year and will vote in this year’s election was born the year Barack Obama was elected president. … The era of politics for an 18-year-old today is chaos, Covid, and corruption. They do not remember the hopium of Obama and the beautiful coalition that we built. That is all ancient history to them.”

A million new voters a year, completely up for grabs. Nobody is inheriting these voters. Neither party owns them. Because as Valencia put it: Demographics have proven not to be destiny — effort and intention matter more.

What I’m hearing

This week I called Matthew Eadie, a reporter at Evanston Now, who has been covering an ICE stop in Evanston, Illinois that has since attracted national attention. On the morning of Sunday, September 20, federal agents stopped a man walking on Howard Street. He turned out to be a US citizen, and not the person they were looking for. Agents reportedly left him on the ground with injuries to his head, neck, and teeth. DHS says he “refused to identify himself. Now the community is demanding answers.”

What’s the community reaction been?

People are really furious. If you remember the fall here — Operation Midway Blitz, the heightened Border Patrol enforcement across the Chicago area, Gregory Bovino parading around the streets with camera crews, the protests at the Broadview processing center — people here were very aware of how impactful it was. It was also happening around the elections, so it was prominent news everywhere and everybody was talking about it.

Over the summer it really died down, and I don’t think people have paid much attention since. ICE has been around all year, but they haven’t been as flashy about it, which has been talked about nationally too. This was a flare-up, and it’s a reminder that it’s still going on even after Bovino left. There have been protests at Broadview this week unrelated to the Evanston incident. So either people are starting to pay more attention as the election gets closer, or the incidents are actually picking up again.

What are the unanswered questions you’re still chasing?

Who were they looking for? Were they actually looking for somebody? That’s been their story — a specific individual — and a lot of people I’ve talked to are really skeptical that’s the case. One witness at the scene told me he felt like they just picked him and drove up on him. Essentially what’s been called a “Kavanaugh stop” — if they’re suspicious of you, they can pretty much find a reason to detain you. And then they dropped it once they realized either that he was a US citizen, which he says he is, or that there were too many people around taking videos for it to be worth their time.

So: What happened in the lead-up, and who specifically were they looking for? 

Afterward, the agents talked to Evanston police — that’s an Evanston police policy; they’re supposed to try to get information from ICE officers. I’ve FOIAed all the body cam footage. I want to see what those conversations were like, and what exactly the agents told police officers when they talked privately.

Part of what we’re trying to do is see the issues that play out from the bottom up rather than the way they get framed nationally. Beyond immigration, is there anything in your coverage that’s shaping local politics more than the national conversation reflects?

The national issues matter to everybody, but they hit communities in completely different ways. Evanston is affluent — college town, Northwestern, wealthier than most suburbs. Affordability is a buzzword that comes up everywhere, but people here talk about it differently.

This is an overwhelmingly progressive Democratic city. Some people in Evanston certainly need it to be more affordable for themselves. But you also have progressive voters saying we want to make Evanston more affordable and more inclusive for other people. Sometimes the affordability conversation gets framed as “life is too expensive for me and we need to make it cheaper for me,” which is true in a lot of places. In somewhere like Evanston, where there are a lot of affluent people who are also very progressive voters, it’s more: If we actually want to say we’re a progressive community, we have to show it. And showing it maybe means building affordable housing.

It’s also a city that passed reparations, which the federal government is challenging right now. That’s going to be a federal court case through the end of this year and probably into next year — the feds challenging the city’s reparations program. That’s something the city prides itself on. Those are big issues here that come out of the national conversation but sound different than how we hear them in national media.

If there’s one thing that Americans can unite behind, it’s anxieties over artificial intelligence.

According to a recent NBC News/Decision Desk poll, 70 percent of adults say they are personally more worried than excited about the technology; it’s basically underwater in every area of society with one notable exception: its use in scientific research and medicine.

According to Dr. Dhruv Khullar, a physician at Weill Cornell Medicine and writer at the New Yorker, that optimism is well-founded.

Khullar uses AI tools in his own practice as a physician and writes about their advancements in the medical field. He says that “right now, at least, there’s a generally optimistic view of how AI will change healthcare going forward.”

AI in medicine is nothing new. In 2024, the Nobel Prize in Chemistry was given to researchers who developed an AI model that enabled breakthroughs in predicting protein’s complex structures. As it becomes more ubiquitous, it’s also emerging as a resource for doctors to get second opinions and as a tool to take notes on patient-doctor interactions.

On top of that, Khullar says, it might help unlock new medicines to treat diseases. 

Khullar joined Today, Explained co-host Sean Rameswaram to discuss the ways that AI is transforming drug discovery and research, as well as what it still can’t do. They also discuss how to think about AI’s potential in healthcare at the same time as top AI researchers are increasingly warning about the technology’s risk.

Below is an excerpt of the conversation, edited for length and clarity. There’s much more in the full podcast, so listen to Today, Explained wherever you get podcasts, including Apple Podcasts, Pandora, and Spotify.

You’ve been thinking about lifesaving AI innovation. Has AI already started saving lives?

I think it has. This is the fastest I’ve ever seen healthcare take up a new technology. Part of that might be that the healthcare system is so messed up that there’s an appetite for some type of change. It’s unaffordable, it’s inaccessible, it’s inconvenient, the quality is uneven. 

There’s at least a hope that AI is going to help with all of these things, and I think it’s already starting to make its way into the healthcare system. Now it’s our responsibility to figure out how to make the most of it without also promulgating some of the downsides.

For those who are unaware, tell us how it’s working its way into the system.

It’s helpful to categorize it into a few buckets, at least for healthcare. 

The most rapid uptake, and I think the place that it’s already starting to make a difference, is on the administrative side. Many people who have gone to a doctor recently might have noticed that there’s an AI scribe that’s taking notes now. And people seem to really like it, at least so far. Doctors are able to look their patients in the eye in the way that they weren’t able to when they were poring over their computers and just typing what the patient was telling them. 

The second big area is patient navigation. When is your next scan? When’s your next appointment? Do I need to take this medication on an empty stomach or not? There’s a huge opportunity for people, let’s say, who are diagnosed with a serious illness like cancer or heart failure to navigate the system more seamlessly. 

The third big area that I’m really excited about is drug discovery. AI is making a huge dent in the early parts of drug discovery. 

The last thing that I’m excited about and that I use every day when I’m in the hospital is [AI] as a second opinion. Now instead of having to get a consult or turn to a textbook, AI can be a very, very helpful clinical decision support. It’s not to say that I never consult someone, of course, but that first pass — I’m not sure what’s going on. What are some recent trials that might influence my decision here? Is there something I’m missing? Is there a test that I should be ordering that I’m not ordering? — all those types of things, they start to help with.

The biggest chunk of this positive view of AI and medicine is drug discovery. Tell me more about how that’s going so far and how it might be going in a year, or five, or 10.

I don’t want to leave this conversation and say that we’ve solved the cure for cancer because of AI, but I do want people to know that at least for the early stages of drug discovery — figuring out the very basic steps of is this an interesting molecule? Does it have potential biological applications? — AI is already very helpful for that. 

They’re trying to take, basically, what is an infinite number of drug-like molecules that are theoretically possible to be drugs and they’re trying to match that to some disease that’s going on inside a human. And human biology is incredibly complex as well. There’s tens of thousands of genes, hundreds of thousands of proteins, trillions of cells, and you have to match a potential molecule to a potential target within a human. 

In the past, a researcher might take years to study a disease or a biological pathway and figure out, okay, this protein seems to be involved. And actually, it might not be involved, it might be involved, it might be disrupted, but not be the causative agent. There’s a lot of uncertainty there, so now AI can take enormous amounts of data, and it can rank various potential targets that seem to be the most likely to be causing a particular disease and give you that short short list of potential targets that may have taken months or years in the past.

So now you know what to attack, but you have to figure out what to attack it with. So you need a molecule, you need to generate a molecule that’s going to fit into that target or disrupt that target in some way. 

Now again, AI can scour vast chemical spaces and give you a list of the things that seem to be most likely to be able to disrupt that target. And then it can actually help you generate that molecule in some cases. 

So now we have the target. Now we have something that we’re going to attack the target with. And then you have to do something called lead optimization, and that’s the third thing. Lead optimization means you have this lead and you have to optimize it.

“I would make the counterargument that healthcare is ripe for disruption and you don’t even need the frontier frontier models to make it a lot better with AI.”

Now we have to figure out, of all the molecules that seem to be good at potentially affecting this target, which ones are able to be absorbed in the body, which ones are going to get to the right tissue. AI models can predict the most likely molecule to give you that Goldilocks set of properties that’s needed for a safe and effective drug.

As we all know from anyone who’s asked AI to write an email for them or Googled a question, these tools have a certain degree of certitude, even though these tools can just be flat out wrong about stuff. Are you guys worried about that in the field?

Absolutely. And so this is why I think the narrative around AI just replacing scientists or doctors or other workers is incorrect. Because you still need a lot of judgment. 

You need to be able to adjudicate the output of these models to figure out what is most promising and what is potentially dangerous. That is going to require wet labs and scientists and reasoning based on prior experience and understanding the context. There’s all sorts of issues around, can you manufacture some of the drugs that are being proposed by these models? So just because it’s set, it dreams something up doesn’t mean you can actually make that thing in the real world.

Some of the drugs that it proposes might actually be toxic in certain ways that it didn’t predict. And of course, then you have to take this thing into clinical trials. You have to recruit l people who are willing to put this medication in their bodies. You have to find out who might benefit, you have to put it through the regulatory process. 

That’s why I said that the first part of drug discovery, where you’re trying to figure out how do we get the right drug candidates, what seems most exciting to test further, that’s going to be really accelerated. That whole second half where we actually have to figure out if it works in human biology, that is still going to be in some respects an analog process.

It doesn’t sound like you’re terribly scared that we are like ceding control of our hospitals, of our research facilities, to AI. This is very much in the sort of assistant bucket?

For now it is. And I think one of the challenges is to maintain our agency as these models become more and more sophisticated. As you lean more on these machines, if you’re a doctor, inevitably some of the skills, the critical thinking, the reasoning that we put into coming up with the diagnosis that we honed over the course of years, that can start to atrophy. 

These are the types of things that we still need to sort through as we’re implementing more and more AI into the healthcare system.

For the people who just say, shut it all down, this is too dangerous, the risk is far too great, we don’t actually need this, this isn’t doing anything for society, would you make a counterargument?

I would make the counterargument that healthcare is ripe for disruption and you don’t even need the frontier frontier models to make it a lot better with AI.

Even if we wanted to “slow the pace of the frontier,” fine. But there are models from a year or two years ago that could be very helpful in the healthcare setting. When we’re talking about slowing the pace of the frontier and the most sophisticated and potentially dangerous models, fine. 

But if we’re talking about shutting down AI and not using it in biotechnology or not using it in scientific research or not using it in clinical care delivery, that’s where I would push back pretty hard.

Who gets to tell the military “no”?

That question has become even more important after the Pentagon pushed AI companies to loosen limits on how their technology may be used in war — including how much human control must remain over autonomous weapons. When Anthropic said no, it lost a major government deal. OpenAI stepped in, embracing the Pentagon’s more flexible standard instead. Now, thanks to new reporting from The Intercept, we’re getting a better look at what those negotiations actually entailed.

The OpenAI-Pentagon agreement included specific language on the type of AI models the DOD would receive: “OpenAI models that are designed for national security use cases and have minimal refusal rates.” That phrase — “minimal refusal rates” — means that the Pentagon is seeking models that rarely, if ever, tell it no. (According to The Intercept, a Justice Department lawyer representing the Pentagon confirmed this language, before backtracking only a few hours later.)

On the surface, that provision might seem reasonable. The military is a strictly hierarchical institution. When your senior gives you an order, you do it. Leaders generally have more battlefield wisdom and experience than their subordinates, so the rank and file can usually assume that their commanders know what they’re doing and follow along. Plus, in the high-stress, time-sensitive environments of most military operations, there is little room for discussion, explanation, or alternative proposals. 

So, if soldiers can’t refuse an order, why would a machine be allowed to? Who is a robot to tell its commander no?

There’s just one issue. Soldiers can refuse an order — when the order is illegal. In fact, they are legally required to. But by introducing technologies with a limited ability to challenge commands, the military may be making it harder to stop war crimes before they are committed.  

Humans have a duty to disobey

As outlined in US domestic, military, and international law, soldiers are only obligated to follow lawful orders, and they have a duty to refuse unlawful ones. 

The Department of Defense’s Law of War Manual is unambiguous: “Each member of the armed services has a duty to: (1) comply with the law of war in good faith; and (2) refuse to comply with clearly illegal orders to commit violations of the law of war.” Obeying an illegal order — as in a directive that violates the Constitution, US laws, international law, or directs one to commit a crime — can expose a service member, as well as their commander, to criminal liability. 

This “duty to disobey” has foundations in early British and American common law, but it achieved international recognition after the Nuremberg Trials against Nazi soldiers after WWII. 

In those trials, many Nazi soldiers argued they were not liable for crimes they committed during the Holocaust because they were simply following their superiors’ orders. Rudolf Hoess, the commander at the largest Nazi concentration camp, Auschwitz, himself explained: “In Germany it was understood that if something went wrong, then the man who gave the orders was responsible. So I didn’t think that I would ever have to answer for it myself.”

These defenses were unequivocally rejected by the court.

The tribunal held that obedience is not without limits, especially when a wartime act would be clearly illegal to the subordinate or any person of “ordinary sense and understanding.” If “I was just following orders” was an absolute defense, everyone involved in a crime would escape accountability, with the exception of the single highest-ranking officer. Instead, the court determined that individual moral and legal responsibility supersedes national or military chains of command for illegal acts.

Obedience is not without limits, especially when a wartime act would be clearly illegal to the subordinate or any person of “ordinary sense and understanding.”

But, when it comes to agentic AI systems or AI-assisted weapons, increasingly being used in place of human soldiers, does the same logic apply? And, perhaps more importantly, should it?

…But do machines?

When you use an AI chatbot, it does not simply reject queries it may not “agree” with. It does not say no to requests it doesn’t “like.” Instead, it follows a set of established protocols, much as a soldier should follow the law of war: It refuses requests that violate its terms of service. 

The military is using AI for far more consequential applications than a layperson using ChatGPT to, say, build a travel itinerary. So, what kind of world is created by the Pentagon’s request for AI that “minimally refuses” its directives, many of which may bear grave ethical, moral, or legal implications?

Rebecca Crootof, an expert on technology law and professor at the University of Richmond School of Law, told Vox that while the language could easily be read as increasing the likelihood of compliance with unlawful orders, or at least reducing the friction around them, “Minimal refusal doesn’t mean no refusal.”

“I both want to say it’s not definitely going to mean compliance with unlawful orders,” Crootof said, “but also acknowledge that identifying and evaluating what constitutes an unlawful order in the moment can be incredibly difficult.”

Even for human soldiers, disobeying illegal orders is neither simple nor easy. To the contrary, it doesn’t even happen that frequently. In fact, some legal scholars have argued that giving illegal directives in and of itself is an act of abuse against subordinates, who are not thoroughly trained in legal nuances, are conditioned to obey, and face prosecution if they make the wrong choice. 

Still, the mere fact that soldiers can disobey is an important constraint on commanders who might let battlefield conditions, or even commands coming from their own superiors, erode their ethics and commitment to the law.

Still, the mere fact that soldiers can disobey is an important constraint on commanders who might let battlefield conditions, or even commands coming from their own superiors, erode their ethics and commitment to the law.

Those who are optimistic about AI often argue that machines could be better at interpreting and applying the law than human soldiers. Machines can be trained on the law itself — drawing on a repository of legal cases, rulings, and battlefield data — which might make them more “knowledgeable” than an average soldier. And machines do not face the same pressures as soldiers facing prosecution for making mistakes, since they can’t exactly be held accountable; a problem to unpack another time. That, in turn, might make it easier for them to refuse illegal orders in the proper conditions. 

So, is the answer as simple as letting the machines decide? Not exactly. “This has been a desire and a drive by technologists for ages,” Crootof said. “What if we just put the law in the system and have it only do things that are lawful? Isn’t that going to solve things?” 

Interpreting and applying the law, though, is not a simple legal-illegal binary. “It’s very context-specific,” Crootof told me, “and honestly, that’s the kind of thing humans generally find difficult and AI decision-making systems are even worse at.” Often, determining the legality of an act comes down to subtle nuances you can’t program for: You’re allowed to target a combatant; you’re not allowed to target a wounded or surrendering combatant. You’re not allowed to target a civilian; you are allowed to target a civilian directly participating in hostilities. The difference between a legal target and an illegal one may come down to body language, clothing choices, or even battlefield intuition. 

As Crootof explained, “These are not the kinds of contextual analyses AI is good for.”

So what’s the solution?

Where human soldiers’ power to disobey would traditionally serve as an important check on the power of military commanders, if an imperfect one, replacing soldiers with machines that don’t have such safeguards removes one last layer of judgment over the use of force. And yet, machines themselves are not always capable of making the judgment calls that go into interpreting the law, so it may not be appropriate to let AI have the final say either.

So what’s the solution? Given the way AI technology has already been thoroughly integrated into the military, Crootof told me it may be somewhere in the middle. 

Rather than programming AI to refuse commands it interprets as illegal, Crootof said, “what it could be designed to do is flag those indeterminate zones, those gray-zone analyses, for human review or for a higher authority.” When a system picks up conflicting signals — an enemy soldier who looks like he may be surrendering, for example — it could prompt the commander to confirm their selection or remind them of applicable law. 

In this way, a check on commanders’ battlefield authority could be preserved without fully ceding judgment over the use of force to the machine. 

We don’t know yet whether OpenAI has equipped its “minimal refusal” models with such capabilities. But as those technologies become increasingly central to military decision-making, the public deserves answers sooner rather than later. 

Nationally, abortion has faded as a campaign issue. Four years after it helped Democrats stave off a red wave, candidates in battleground races are focusing their attention on cost-of-living issues, immigration, and policing. As of mid-August, abortion was mentioned in less than 5 percent of Democratic Senate ads, a key difference from the past two cycles, according to the Wesleyan Media Project.

Maine is different. As of September 16, Democratic advertisers had spent roughly 15 percent of their $40.8 million broadcast TV and streaming dollars on ads focused on abortion rights, according to Nate Schwartz from AdImpact. It’s one of the biggest topics there, he told me, after ads about President Donald Trump and healthcare. Planned Parenthood’s political arm is also making its biggest ad buy of 2026 against Susan Collins, at $600,000, and WinSenate, a super PAC aligned with Senate Democratic leadership, has spent 70 percent of its TV, streaming, and digital money in Maine on abortion-focused ads.

The target is Collins, the only Republican senator up for reelection in a state Trump lost in 2024, and one of Democrats’ best chances to flip a seat. Seeking her sixth term, Collins is running in a year when Democrats lead the generic congressional ballot by about 9 points. It’s also the first time she will face voters since Roe v. Wade was overturned — a decision Justice Brett Kavanaugh helped deliver after Collins, who has long described herself as “pro-choice,” cast the pivotal vote to confirm him in 2018. 

For the abortion rights movement, this is the race that has to work. Groups are pouring their biggest investments of the cycle into a state where fewer than half a percent of voters call abortion their top issue, in a year when Democrats have opportunities to win nearly everywhere. If a blue wave arrives and lifts every Democrat except the one running to avenge Roe v. Wade and remove the senator who famously sealed its fate, the organizations will have a harder time telling donors the issue still moves voters, and a harder time arguing that Republicans will pay a political price — even in blue states — for votes that restrict access to abortion.

“We know that abortion is the number one negative in Maine voters’ mind,” Lisa Margulies, a Maine spokesperson for Planned Parenthood Votes, told me. “It’s really the No. 1 dealbreaker.”

The first job for Democrats and abortion rights advocates is convincing Mainers the issue is still a pressing one. Seventy-two percent of Maine adults say abortion should be legal “in all or most” circumstances, and since Democrats took control of Maine’s government in 2018, the state has expanded who can provide abortions, required and subsidized insurance coverage for the procedure, eliminated most patient cost-sharing, loosened restrictions on care after viability, and shielded patients and providers from prosecution and lawsuits.

Collins has made that her defense. The Dobbs v. Jackson decision, she said in June, “has not had an impact on the state of Maine in that Maine actually expanded its law.” Advocates call that the central misunderstanding. 

“There’s this perception — and Sen. Collins herself has said it — that somehow Maine is immune, and that’s just not true,” Margulies said, pointing to their state-level efforts to keep access intact and to federal Medicaid cuts that have already hit Maine providers.

Also muddying the waters is the senator’s Democratic opponent. Collins is facing off against Troy Jackson, a longtime fixture in Maine politics who rose from the state House to the Senate presidency but fell short in bids for Congress in 2014 and governor this year. Early in his career, Jackson said that abortion “should never be legal” and that women should be required to view an ultrasound if they want to end a pregnancy. He has since renounced those positions and has earned 100 percent ratings from Planned Parenthood since 2017. (Jackson replaced Graham Platner, the original Democratic Party nominee, who quit the race after being accused of rape, an allegation he denies.)

While Collins prevailed for reelection in 2020 despite voting to confirm Kavanaugh, Democrats say her decision is much more salient for voters now that Americans live in a post-Roe world, one where more than 15 states have near-total abortion bans. 

If Collins loses, Democrats and abortion rights groups will say her Kavanaugh vote is no doubt partly why. A win for Collins would be harder to read, though the explanations that leave out abortion are already being prepared, including a messy nomination process that produced a candidate fewer voters were excited about, and a “unicorn” incumbent with decades of personal relationships across a small state and a long record of bringing federal money back home.

The abortion rights bet hinges on two things: that the Kavanaugh vote cost Collins something on top of what an unpopular president and rising gas prices already have, and that voters will judge Jackson by the positions he’s renounced rather than the ones he once held. 

The vote Collins won’t apologize for

When Collins voted to confirm Kavanaugh, she said she believed his assurances that he considered Roe v. Wade settled precedent. She had previously said that she would not vote for any nominee who expressed “hostility” toward the 1973 decision that established the constitutional right to an abortion.

As many observed at the time, Kavanaugh’s record gave strong signals that he would in fact vote to overturn Roe v. Wade. In her floor speech announcing her vote, Collins dismissed those warnings, suggesting opponents had been whipped into a “frenzy” by advocacy groups. When he ultimately did side with the conservative majority in Dobbs, Collins issued a statement calling the ruling “ill-considered.”

Four years later, though, Collins is doubling down on her vote for Kavanaugh, telling local media this summer that she does “not regret” it. “Given what subsequently happened to Roe, I don’t think that simply fades into the background,” Robert Glover, a political scientist at the University of Maine, told me.

Democrats are zeroing in on the fact that Collins has continued to confirm anti-abortion federal judges, even as recently as this month. Last fall she voted to confirm Joshua Dunlap, a Maine attorney, to the US Court of Appeals for the First Circuit over objections from 20 Maine organizations citing his record on abortion and marriage equality. 

“Susan Collins says she has no regrets about confirming the judges who overturned Roe v. Wade, and has proved it by continuing to vote for Trump’s right-wing judges,” Maine Democratic Party spokesperson Misha Linnehan told me in an emailed statement. “Now, her approvals are underwater, and she’s trailing with women voters because Maine women have had enough of being betrayed.” 

George Hill, the president and CEO of Maine Family Planning, told me Collins “defied” them with her vote for Kavanaugh, causing “irreparable harm” across the country. He emphasized how vulnerable Mainers are if Republicans retake control of state government.

Blake Kernen, a spokesperson for Collins’s campaign, dismissed the criticism about anti-abortion judges, emphasizing that her pro-choice boss also voted for 188 judges nominated by President Joe Biden, including Julie Rikelman, who served as litigation director of the Center for Reproductive Rights. Collins, she noted, is the only Republican who voted to confirm all three of the court’s liberal justices, and she voted against Amy Coney Barrett.

Margulies rejects this accounting. “We deserve better than, ‘Well, she’s not as bad,’” she said. “When it comes to reproductive rights, she’s not a moderate anymore.”

While her judicial votes are facing scrutiny, Collins has also spent decades building up a brand as a Republican willing to buck her party on abortion when the issue comes up in the Senate. 

In May 2025, Planned Parenthood Maine Action Fund published a memo online titled, “Senator Susan Collins’ long history of supporting Planned Parenthood and access to essential health care.” Among other things, the memo highlighted Collins for speaking out against proposed Medicaid cuts, for co-sponsoring a bill in 2022 to codify Roe v. Wade, and for opposing a 20-week abortion ban bill in the Senate. This past summer, the memo was quietly taken down from their website.

Asked about the removed page, Margulies didn’t say when or why it was taken down, but pointed to what she described as a sharp shift in Collins’s record over the past year, particularly on judicial confirmations. “There are times when she’s been true to her word and put Mainers ahead of politics, but she closed that chapter this year,” she told me.

“Because she is a Republican, the national groups continue to twist themselves into a pretzel to support these male candidates with terrible records,” Kernen said, referring to Platner and Jackson. 

So far, Mainers appear to have split Collins in two in their minds: There’s the dedicated public servant who puts her state first on one side, and the partisan Republican who sometimes disappoints them on the other. A University of New Hampshire poll over the summer found Collins rated far better on bringing federal dollars home to Maine than on her Supreme Court confirmation votes. Maine voters, Glover said, “have historically been willing to separate Collins from the national Republican Party and evaluate her through a broader lens of incumbency, independence, and constituent service.” 

The turnout math

Even if Democrats can win the argument on abortion around Collins, it’s an open question how much it will move votes. “Right now, I feel quite confident that affordability, healthcare, and good jobs trump abortion access by a long shot,” Dan Shea, a political scientist at Colby College, told me.

But abortion rights advocates believe the ranking obscures how the issue actually works. Mini Timmaraju, president of Reproductive Freedom for All, said her group has spent years testing how to talk to voters about abortion, consistently finding that it overperforms as an attack. “If you go to a voter in Maine and you say, what are your top three or five issues, they may not name abortion,” she told me. “But if you say to them, Susan Collins took the following vote — and one of them is Brett Kavanaugh, which led to overturning Roe — it is extremely salient, and it moves that voter.”

The Kavanaugh vote works, she argued, as a character verdict — “an indicator that you’re willing to side with extremism over your own constituents.” That, she said, is the piece that’s missing in the analysis of what voters say they care about. “It is one of the most salient attacks in Maine and in a lot of other parts of the country.”

Ashley Kirzinger, a pollster at KFF based in Maine, said priority rankings are a weak guide anyway, noting “the economy” means different things to different voters, and issues compete with candidate dynamics and party loyalty. What has changed since Roe’s overturn, she told me, is who the single-issue abortion voter is.

“Before the Dobbs decision, abortion was a motivating issue for Republican conservative voters,” Kirzinger said. “Now what we’ve seen is a complete shift.” Those voters are more likely to be Democrats, younger women and women of color, and include more people who don’t reliably show up in midterms.

Abortion messaging, therefore, is doing two jobs. The first is turnout. Kirzinger said the Kavanaugh vote is aimed at Democratic women “who are not as enthusiastic about Jackson as a candidate” and who “may be more motivated to turn out as a vote against Collins’ record on abortion rather than for Jackson.” In a tight race inside the margin of error, a couple of points of turnout could make the difference.

The second job is persuasion, and the target is older women, who reliably vote. Women made up nearly 60 percent of Maine’s turnout in 2020, and Collins won them by 3 points. What Democrats need from them is a switch, which is why many of the abortion ads are built around regret. One Planned Parenthood ad features a woman who voted for Collins and “will not make the same mistake twice”; WinSenate’s features a voter who “felt [Collins] was the better choice” and doesn’t anymore. 

Jackson’s change of heart

Complicating the race is Collins’ opponent, Troy Jackson, who is in some ways a mirror image of her own political persona: a Democrat with a history of culturally right-leaning positions to appeal to voters in the other party.

Jackson, who started his political career in the rural and fairly conservative county of Aroostook, used to earn 100 percent ratings from Maine Right to Life. In 2010, as CNN’s KFILE reported, when he filled out the group’s candidate questionnaire saying he thought abortion should never be legal, he declined to say whether there should be exceptions even to protect the life of the mother, for rape or incest, or because of fetal disability. Maine Right to Life, which has not issued an endorsement in the Senate race, did not return requests for comment.

Jackson told MS NOW he evolved on the issue thanks to “a lot of conversations over time” with women, doctors, and community members. “It became more and more clear to me that abortion access was healthcare,” he said.

In 2023, while serving as Senate president, he co-sponsored a bill allowing abortions after fetal viability when a physician deems them necessary, one of the most expansive such laws in the country. In 2024, he shepherded through the Senate a shield law protecting Maine providers and patients of reproductive and gender-affirming care from other states’ bans, despite bomb threats that emptied the statehouse and a letter from 16 Republican attorneys general threatening to sue. “Troy stood up for and never wavered in his commitment to reproductive rights during some really harrowing moments,” Margulies said, “when it would have been a lot easier to step down.”

“I’m rock solid on this now, and Susan has been the opposite,” Jackson said in an interview with Maine Public Radio this summer. Kernan, Collins’s spokesperson, has charged Jackson with flip-flopping for “political convenience.”

The national environment is looking strong for Democrats, and Jackson has been narrowly ahead in recent polls, though all have been within the margin of error. A CNN poll released this month found nearly two-thirds of Mainers feel negatively about the Dobbs decision, including about half saying it makes them angry. A Fox News poll in August found 55 percent worried Collins votes with Trump too often. 

Whatever happens, the explanations are ready. If Collins loses, abortion rights groups will cast it as proof that the issue still moves voters even where it doesn’t top their lists, and as a warning to all Republicans running in 2028. Timmaraju previewed the argument in our conversation: A loss would signal “a crack in the armor” of so-called moderates, and show that voters “are not going to tolerate” incumbents complicit with an extremist administration. 

If Collins wins, the groups that spent millions on abortion ads are unlikely to conclude that doing so was a mistake. Kirzinger pointed to the turmoil that followed the Democratic primary and the late swap of Platner for Jackson, which she said benefited Collins “in a way that is hard to capture in polls.” Glover, the political scientist, noted the things Mainers have always weighed alongside her votes, including decades of relationships, the Senate Appropriations gavel, and federal money. Even Timmaraju, the abortion rights leader, offered the same hedge in advance: A Collins win, she told me, “says more about Susan Collins and Maine than it does about 2028.”

A win wouldn’t show that judicial votes carry no consequences, Glover added. But if Collins holds substantial support among voters who support abortion rights despite the Kavanaugh vote and the fall of Roe, “that would certainly suggest limits” for retrospective accountability for downstream consequences of Supreme Court confirmations.

Though the race is supposed to clarify whether voters will punish elected officials for taking actions that restrict abortion access, it probably can’t — at least not definitively. It will decide, though, who holds the seat, and reproductive rights leaders have been clear that without a Democratic Senate, nothing they want gets a vote.

Mike Huckabee, the US ambassador to Israel, is a Christian Zionist who believes that Israel is biblically entitled to all of the Holy Land. He’s pretty much the last person you’d expect to condemn Israeli conduct in the West Bank, where settlers have spent decades using a combination of construction, coercion, and outright violence to push Palestinians off their own land.

Yet earlier this month, Huckabee went out of his way to denounce the behavior of some settlers — describing them as “band of terrorists” who had committed “horrible activities of criminal nature.”

Huckabee’s surprising outrage stems, in part, from their recent choice of victims. Last month, settler extremists besieged three Palestinian homes in the village of Qusra, including one belonging to American citizen Lou Ridi. About two weeks later, masked settlers beat up a NBC News crew while they attempted to interview Aida Ahmed Abdullah, a Palestinian who had herself had been forced from her home by settler aggression.

But Huckabee’s outrage reflects something deeper. Settler terrorism, once relatively uncommon, has exploded in severity in recent years — with wink-nudge support from Prime Minister Benjamin Netanyahu’s government.

The violence is so egregious, and so indefensible, that retired Israeli generals have compared the settlers to Nazis and demanded action to stop them. In mid-September, the INSS — Israel’s top national security think tank — issued a strategic alert warning that “Jewish terrorism” had become a national crisis and even an existential threat to Israel’s future.

“As history teaches us, democracies that fail to restrain radical forces within them do not survive,” the alert argues.

In Washington, settler terrorism against Palestinians has flipped the Israel debate on its head. Congressional Democrats, who have long been divided over the Gaza war, have come together to support legislation that would impose sanctions on Israeli politicians and corporations implicated in the violence. The Trump administration, the most pro-settlement White House in American history, has very conspicuously stood aside as Canada and European nations implemented sanctions of their own.

For now, the settlers are mostly getting away with it — because the government wants them to.

For now, the settlers are mostly getting away with it — because the government wants them to. While basically all Israeli governments have allowed settlements to grow, the current Netanyahu government includes coalition partners unusually aligned with their most radical goals: the ethnic cleansing of the West Bank. So long as they see settler terrorism as a useful tactic toward that end, and have the ability to stymie preventive measures and block accountability, the attacks will continue.

But if Netanyahu loses Israel’s election in October, and Democrats retake the White House in 2028, things might change dramatically. The political stars could align for a serious effort to change what Israel is doing in the West Bank — not just attempting to merely slow settler violence, but also challenge the broader apparatus of occupation that makes it possible.

In “conversations that I’ve had with congressional Democrats and their staff, this [settler violence] is the No. 1 issue that comes up, the thing they’re angriest about,” said Michael Koplow, the chief policy officer at the center-left Israel Policy Forum. “If you have a Democrat in the White House, this is something the Israelis are going to have to deal with — whether they want to or not.”

The rise of settler terrorism

Under international law, Israeli settlement in the West Bank is illegal: Countries are not permitted to move their own civilians into territory they occupied after a war. However, under successive Israeli governments — including the more peace-minded ones — the settlements have continued to grow. Many of those closest to the border function as suburbs of major Israeli cities; some further east, like Ariel, look like small cities in their own right. During the era when a negotiated solution looked possible in the near term, their growth was often seen as part of a bid to dictate the eventual borders of a Palestinian state — or blow up negotiations entirely — with “facts on the ground” that would be hard to undo.

While individual settlers live there for all sorts of reasons — for many, the housing costs are cheaper — the political movement overall exists to seize land from Palestinians and slam the door on the possibility of a two-state solution to the conflict. 

On a macro level, the physical placement of the settlements cut Palestinian communities off from each other, making it hard — basically impossible — to construct a contiguous Palestinian state so long as they remain standing. On a micro level, settlements are often placed in ways designed to push Palestinians off their land — by, for example, putting Palestinian-owned land into no-go zones enforced by the Israeli military. And politically, the population growth in settlements creates a powerful constituency in Israeli politics whose entire way of life depends on preventing a peace agreement that would require settlement evacuation.

Settler violence is the literal bleeding edge of this overall strategy. Violent settlers live in ramshackle communities, tiny villages on hilltops or herding outposts encroaching on Palestinian farmland, that are often illegal even under domestic Israeli law. The people living there are the hardest of the hardcore settlers: fanatically devoted to the idea that the land is theirs by divine right, and that they have every right to use force to push their Palestinians neighbors out.

Their tactics include everything from killing Palestinian livestock to cutting off their water supply to burning down their homes to outright beating and killing them. The violence is autonomous, in that individual settlers decide on what to do, but enabled by the overall strategic plan of the settlement movement.

“There’s no hierarchy deciding that this [specific] house will be attacked. The decision is that this area is an area of all our interest, so we need to subsidize herding farms there. And those bring violence, which lead to displacement” of the local Palestinians, said Yehuda Shaul, co-director of the Israeli think tank Ofek and a leading observer of the settlement movement.

Settler terrorism is not new, but it used to be comparatively rare. In 2012, Natan Sachs — a leading expert on Israeli politics at the Middle East Institute think tank — wrote an essay warning that the growth of the settlements, and the radicalization of some settlers, were creating a dangerous new violent movement of settlers.  

“Back then, it was a fringe thing among settlers,” Sachs tells me. “It has become much more mainstream.”

Israel’s current government, which took power in late 2022 and is by far the most radical in Israeli history, bears enormous responsibility.

“This particular government is enabling settler violence beyond anything we’ve ever seen,” said Yousef Munayyer, the director of the Israel/Palestine program at the Arab Center think tank in DC.

Minister of Finance Bezalel Smotrich, a settler living in a reportedly illegal home and a longtime advocate of ethnically cleansing West Bank Palestinians, is currently in charge of civilian administration in the West Bank. In effect, this allows him to use state resources to orchestrate a historic expansion of settlements across the West Bank — including the illegal outposts that many violent settlers call home. 

“This is his life’s work,” said Mairav Zonszein, a senior analyst on Israel-Palestine at the International Crisis Group. “He’s completely dedicated to this and he knows exactly what he’s doing.”

Minister of National Security Itamar Ben Gvir, also a settler, is if anything more extreme: For years, he hung a portrait of Baruch Goldstein, a terrorist who massacred 29 Palestinians at a holy site in Hebron, in his living room. Ben Gvir has, among other things, instructed the police forces under his authority not to arrest violent settlers and given thousands of weapons to settlers themselves.

In theory, the Israeli military and intelligence establishment should be a check on the extremists. Settler violence is not only illegal, but also a national security threat. Settler violence prompts Palestinian retaliation and, at times, even targets Israeli soldiers and police themselves. It can lead to international incidents, like those seen recently, that undermine Israel’s support with key allies abroad. The Israeli military is the ultimate authority in the West Bank; it could act on these concerns if it so chose.

In practice, however, they’re of little help. Israeli government data shows that, between 2005 and 2025, 94 percent of all investigations into settler violence fail to yield an indictment. The military will facilitate illegal settlements’ connection to the electrical grid. And its ranks are increasingly made up of “national religious” Jewish Israelis, a demographic that’s theologically on board with the settlement enterprise, versus more secular Israelis who primarily see the conflict through a security lens.

And after October 7, the Army dealt with a manpower crisis by significantly increasing the number of “regional defense units” — essentially bands of armed settlers that can be tasked with local security in the West Bank in lieu of regular soldiers. It emboldens violence in exactly the way you’d think.

So while the recent spike in settler terrorism can fairly be blamed on the Netanyahu government, the phenomenon itself is an outgrowth of the structural settlement growth Israel has condoned or actively pursued for decades.

“It’s all one set of policies,” said Ilan Goldenberg, senior vice president and chief policy officer at J Street, an American lobby that favors a two-state solution to the conflict.

How the violence could lead to a deeper break between America and Israel

Other than Israel’s government itself, the most important player in all this is the American government. As Israel’s chief diplomatic ally and military supplier, it has immense leverage that could be used to affect the West Bank (or anywhere else). Thus, changing American politics on Israel could quite directly produce a transformation on the ground.

Since October 7, Gaza has dominated American discourse over Israel — and it is arguably the most divisive issue in both the Democratic and Republican parties. Many Democrats (and a majority of rank-and-file voters) call Israel’s actions in Gaza genocide, a label others in the party strongly reject. While the Trump administration and most elected Republicans have stood behind Netanyahu, a faction — including Vice President JD Vance, Tucker Carlson, and a significant chunk of younger GOP voters — have grown increasingly critical of American military support for Israel’s actions.

But settler violence, specifically, might as well be another political universe. 

On the Democratic side, there is basically unanimous condemnation of what’s going on in the West Bank. Even Democratic Majority for Israel (DMFI), a leading pro-Israel advocacy group, believes that Israel is erring here.

“DMFI unequivocally condemns attacks by violent Israeli settlers against Palestinians in the West Bank. Israel’s government must move beyond just words and take definitive action to protect civilians, and ensure that those responsible are held accountable,” Brian Romick, DMFI’s president, told me.

Last year, US Rep. Jerry Nadler (D-NY) — an influential progressive pro-Israel voice in Congress, where he’s long represented some of the nation’s largest Jewish communities — co-authored the West Bank Violence Protection Act, legislation that would sanction violent settlers and their enablers. The law would effectively codify an executive order issued by President Joe Biden in 2024, but scrapped by President Donald Trump upon taking office. 

The key question going forward is what this political dynamic means for policy.

More than 170 House Democrats and 45 senators have cosponsored it, the vast majority of Democrats in both chambers. Goldenberg, whose J Street outfit has been a leading voice supporting the law, says rallying support for it has been easy.

By contrast, no Republicans have cosponsored the bill. When you combine that with Trump reversing the Biden order, and his broader enabling of settlement expansion, you might think that the administration would be willing to take Israel’s side in the argument over settler violence.

Yet that’s not been the case — and it’s not just Huckabee freelancing. Earlier this month, when Canada, the UK, and France imposed significant sanctions targeting settlements, Secretary of State Marco Rubio pointedly refused to condemn these allied policies, saying that neither he nor Trump had any comment on their decisions.

“Obviously, we’re not going to do what the UK did. [But] we share the goal of stability: we don’t to see an uptick in violence or conflict or tensions in the West Bank,” he told reporters on September 8.

Settler terrorism, then, has a remarkable effect on both parties: Democrats in Congress go from fractious on Israel to united, and the Trump administration goes from Israel’s chief defender to a tacit enabler of allied pressure campaigns against it. Why?

I put this question to all the experts I spoke with, and everyone offered a basically similar explanation: that settler violence is morally and politically straightforward. Even the most hardline pro-Israel advocate, ones who rationalize mass civilian killing in Gaza as necessary to destroy Hamas and restore Israel’s security, blanch at videos of armed settlers beating up on defenseless Palestinians, which cannot be justified on the same grounds. Virtually nobody in American politics wants to be in a position of having to defend this kind of thing, and many of the most influential pro-Israel voices on the American side do not share the settlers’ extremist ideology.

“There aren’t two sides…it’s just one side terrorizing the other,” Israel Policy Forum’s Koplow said. “There are literally WhatsApp groups and Signal groups run by the Israelis perpetrating the violence where they themselves boast about it.”

On the White House side, there are also Trump priorities that the violence complicates. In his comments about the allied sanctions, Rubio said instability in the West Bank “has implications as well on our ability to make progress on Gaza, on the 20-point plan, and the Board of Peace that we’re very committed to.” However serious you think these initiatives are — there are very good reasons for skepticism — the White House seems to at least be publicly concerned about violence derailing it.

The key question going forward is what this political dynamic means for policy. While the Trump administration may not be willing to go out of their way to defend settler violence, they’re not exactly going out of their way to try and shut it down either. Huckabee’s comments about “terrorists” are symbolically significant, but the reversal of the Biden executive order speaks about the limits of this stance for actual policy. 

Democrats, by contrast, are clearly animated by the issue and willing to act on it. The question is what exactly those actions entail — especially if they retake Congress in 2026 and the White House in 2028.

Here, there’s real disagreement between the experts. 

Skeptics suggest that there may be less substance here than it appears. Targeted sanctions against some violent settlers might matter on the margins, but unless they go after the settlement enterprise and US-Israel relationship more broadly they’re leaving the root cause of the problem intact. In the worst case, a focus on settler violence specifically might even be counterproductive — functioning as a kind of pressure release valve that allows Democrats to satiate their increasingly Israel-critical base without actually doing much that matters.

Such a focus “creates, in a lot of ways, more problems than it solves,” the Arab Center’s Munayyer said. “It reinforces this idea politically that it’s too risky politically to hold accountable the actual decision-makers in the Israeli government, and the real processes at play here — which are driven by the state.”

Others see this as a misread of the political dynamics — one that Sachs of the Middle East Institute called “too cynical by half.” On this line of thinking, targeted action on settler violence is just morally necessary: the violence is really terrible, and if the United States can curtail it even on the margins it matters.

Moreover, they argue, something like sanctions on West Bank settlers opens the door to broader action: creating a real precedent for punishing bad actors in Israel that could create a kind of ratchet dynamic for future action taking on the deeper Israeli policies behind settler violence.

It’s hard to say in advance which of these views are right. Experts in both camps say there’s a lot of uncertainty here: we don’t know what will happen this November, let alone who the Democratic 2028 nominee will be or whether they’ll win. A Democratic trifecta with President Alexandria Ocasio-Cortez would likely be quite different than, say, a divided government with President Mark Kelly.

But either way, how Democrats handle the settler violence question could well be an indicator of where the party is going to go on the US-Israel relationship more broadly — and, by extension, the future for Israelis and Palestinians alike. 

When we think about the stakes in healthcare, we tend to think of them in terms of life and death, being sick or being healthy. The conversation is about your physical well-being and maybe your mental health — and it pretty much stops there.

But health affects so much more than your body’s biology or even just you personally, as a new survey from Gallup and Pivotal, a philanthropic group founded by Melinda French Gates, shows. That’s especially true for women, who frequently report negative experiences with the health system and ongoing struggles to get their concerns taken seriously; poor health affects the trajectory of their careers and their ability to be engaged with their families, friends, and wider community.

It starts at the doctor’s office or hospital: 51 percent of women say they have had a negative healthcare experience in the past five years, versus 39 percent of men. About one in three said that a doctor did not take them seriously when they said they were feeling pain or other symptoms, versus 21 percent of men, and 33 percent said that they either were misdiagnosed or did not receive a diagnosis. The cost of paying for care is another deterrent: 32 percent of women said they have skipped or delayed care in the past year because of the cost, compared to 25 percent of men.

“To see that half of women in the United States who sought care in the last five years came out feeling like they hadn’t gotten the help or care they needed, that’s frankly staggering,” French Gates told me in an email. “It says a lot about how little we’ve invested in women’s health and how much we expect them to just push through pain or accept uncertainty.”

And the effects of those experiences, the failure to receive proper treatment, ripples out into the rest of the women’s lives. More than half of US women, 54 percent, said they had experienced at least one health-related impact on their family life or relationships; 42 percent said their health had led to a negative effect on their career or work. It also affects their community: 32 percent said their health had affected their ability to participate in civic life.

The toll is not only physical but mental: 48 percent of women said they’d dealt with a mental health condition like depression or anxiety. And socioeconomic disparities abound. Black and Hispanic women are more likely to report being in poor or fair health. Women who have lower incomes are more likely to face career struggles because of their health than those who make more.

I had the opportunity to interview French Gates, who has made women’s health a priority in her work, over email about the survey; its findings on community engagement, menopause, caregiving, and more; and what could be done to address the problems that it reveals. Our exchange is below, lightly edited for clarity and length.

Which of these survey findings surprised you most and why?

The number of women [54 percent] who said they experienced a mental health challenge during or after pregnancy should stop us all in our tracks. And the fact that less than one in three women who sought care found it helpful is really concerning. It’s why I feel so strongly about integrating women’s mental and physical healthcare, especially during pregnancy and postpartum. Mental and physical health are so linked together.

Too much of the burden is placed on women to get answers. All of us deserve so much better.

Nearly one-third of women said in the survey that health problems limited their ability to be involved in their community. What does that look like in practice, and what do communities lose when women can’t fully participate?

Women are so often the backbones of their communities, whether they’re rallying their neighbors around a cause, leading the PTA, or serving in elected office. Without having women — and their ideas and voices — in those roles, we’re missing half of the conversation. Issues that affect women and families are less likely to get on the agenda. They’re less likely to be solved.

When a woman has a health challenge, and she has to spend so much of her time and resources chasing down answers, that makes it harder for her to show up in those spaces. We saw that in the survey: 30 percent of women said their health limited their ability to be involved in their communities. And when women couldn’t get the care they needed, that percentage doubled.

Women’s health is often talked about as a personal issue, but this data is helping us prove that it’s also an economic issue, a family issue, a community issue — it affects all of society.

Caregiving and family responsibilities prevent many women from getting medical care, in particular Hispanic and Black women. That seems to create a vicious cycle: Women are often so busy being the caregivers in their families that they neglect their own health. Eventually, they are the ones who will need care.

What do you think can be done to break that pattern? Is there a philanthropic solution? What public policy changes could be made?

Women are often taught to put themselves last and minimize their own needs. They don’t always have time to care for themselves because they’re so busy caring for others. And they might not be able to afford childcare or home health support for their aging parents. Our survey showed one in six women delayed healthcare because of their responsibilities caring for their family members. For Black women it’s one in five, and Hispanic women one in four. Too many women are facing an uphill road to even just get to a doctor’s office.

It’s possible to make it easier for people to balance healthcare and family responsibilities — but it’s going to take action from many angles. That’s exactly why I invest in both women’s health and in caregiving, and I’d encourage more philanthropists to do the same. Businesses can also support employees with good paid leave policies and reliable but flexible schedules.

But for women everywhere to benefit, we need government leaders to prioritize increasing access to affordable childcare, home health supports, and paid family medical leave. These are policies that have the support of large majorities of Americans because these are issues that affect everyone. So there’s a real opportunity here to offer meaningful help to families.

I am struck by our collective failure on menopause: Women report that they aren’t aware of menopause symptoms until they experience them on their own, that they struggle to get their doctor to take menopause symptoms seriously, and that their menopause experience has held them back in their career. What are some short-term and long-term steps we can take to address those problems?

“Right now, women need better care at critical moments in their lives.”

If more than a third of women only learn about menopause symptoms when they go through them, the system is clearly not working. Women need to learn about menopause from their doctors before they experience it. But doctors aren’t getting the education and training they need on menopause while they are in school. One doctor told me she got just one lecture on menopause across all her years in medical school. One said it was mentioned one time. They are seeking out better training but it can be hard to find. That’s why Pivotal is partnering with the Menopause Society to expand physician training on menopause so doctors, nurses, and other healthcare providers can access the information they’re looking for to better help their patients.

Because menopause isn’t just about being uncomfortable; it can also bring anxiety, heart palpitations, insomnia — significant symptoms that can really hold women back. More than a quarter of women who’ve experienced menopause or perimenopause said it made it harder to live up to their full potential at work. One in 10 had to take leave or leave their jobs entirely. That just shouldn’t be. Women shouldn’t have to step back right at the moment when they could be stepping into leadership roles.

The most immediate steps can come from policymakers, especially at the state level. They can include menopause care in insurance and Medicaid coverage and offer workplace protections to women in menopause the way they do for pregnancy, so women can get the care they need without losing their jobs. I’m also focused on longer-term solutions, like expanding training for healthcare providers and investing in research to accelerate treatment options.  

Women are at the forefront of a lot of wellness trends these days and yet, as the survey reveals, women also report encountering a ton of health misinformation on social media. Why do you think women are so often the targets of this kind of content? What can be done to point women toward and build their trust in sound medical information?

One in three women said she left an appointment without a diagnosis, with her pain being dismissed, or with no answers or next steps. You can understand why those women would start looking for information online. And because women’s health has been underfunded and under-researched, there just isn’t as much reliable information out there as there should be, which leaves a lot of room for people to come in with misinformation. Investing in credible scientific research to fill that vacuum will give both women and their doctors more information going into their appointments.

In terms of your own organization’s work, which investments are you making that you think can make progress on the underlying problems revealed in this survey?

We need to address the right now and we need to address the longer term. Right now, women need better care at critical moments in their lives.

One way we’re investing in that at Pivotal is by making it easier for women to get the mental healthcare they need during and after pregnancy. So we are investing in integrated models that put mental and physical healthcare under one roof. I visited a health clinic in Alabama earlier this year where they identify issues like depression and anxiety right at the outset and then help the mom all the way through her pregnancy care, all in the same place. That model is something we strongly believe in.

Better care for menopause will require better training, which we are thinking about in both the immediate and long term. I mentioned that we’re making more menopause training available to doctors who are currently practicing. That is something I’m really excited about because I’ve heard from physicians who really want to better support their patients and are looking for better information so they can do that. And longer-term, menopause needs to be a much bigger part of the curriculum at medical schools moving forward.

Another longer-term focus of ours is trying to make up for decades of lost time in research. We are funding medical research focused on how diseases specifically impact women, in partnership with an innovative organization called Wellcome Leap. They are working to accelerate progress in women’s health on everything from autoimmune diseases to cardiovascular disease in years, not decades, which is how long health breakthroughs usually take. That’s the kind of urgency women need and deserve.

A decade ago, insulin became the poster child for America’s out-of-control drug prices. The 2017 death of Alec Smith, who was rationing his insulin before he died, and repeated price hikes for a 100-year-old medication (the patent for which was originally sold for $1 to a university to prevent it from being exploited commercially) sparked widespread outrage that a lifesaving medication could be unaffordable to the people who depend on it to survive.

And then politicians actually acted. Starting with a Colorado law passed in 2019, states began passing laws capping the out-of-pocket costs for many people who take insulin; 27 states (plus Washington, DC) now have such a policy. Then in 2022, Congress included a provision in the Inflation Reduction Act that established a $35 per month cap on insulin prescriptions for people on Medicare. Given the national response — drugmakers also made splashy public announcements about reducing the cash price that uninsured people pay — you would be forgiven if you thought this was the one health care cost crisis that the US had genuinely managed to solve.

But the diabetes community has a dire warning: It’s not over. Millions of people continue to struggle to pay for their insulin and having their long-term health put at risk because of, in the words of one advocate who put it bluntly, pharmaceutical greed.

“There has been so much attention and promise to action around the insulin price crisis. … There has been so much action,” said Shaina Kasper, executive director of T1 International, a global nonprofit that advocates for people with diabetes. “And shockingly, frustratingly, critically, the data from our 2024 survey showed that the rationing rates were worse than ever. Ultimately we’re facing such immense and monstrous corporate greed. … When we put a Band-Aid on one hole, another one’s going to find a way to pop out and keep profits high.”

Fixing the insulin crisis will require a multifaceted approach: drug pricing reform, insurance mandates, and possibly even public manufacturing. Out-of-pocket caps, Kasper said, were always a temporary fix for a very broken system. There is still a lot more work to do.

Why so many Americans still struggle to afford insulin

Both national and local surveys indicate that the policy changes of the past decade have not had the desired effect.

A study published last year from researchers at the Yale Diabetes Center found that people treated there were rationing insulin at the same rate in 2024 as they were in 2017; about one in four people reported some rationing due to the cost of the drug in both years. When you added in the people who rationed because of insurance delays or supply limits, more than one in three people — 38 percent — were still having trouble keeping this lifesaving medication on hand in 2024. A 2024 T1 International survey likewise found that nationwide, 34 percent of the Americans who use insulin reported rationing the medicine. (About 8.4 million people in the US rely on insulin to survive.)

Dr. Kasia Lipska, an endocrinologist and researcher who treats patients at the Yale Diabetes Center and who co-authored that study, was surprised by their findings.

“I was surprised because I’ve been looking at the policy and advocating for policy change; you think things are moving, they should be really improving,” Lipska said. “But if I talked to the nurses and our staff at the Yale Diabetes Center, they told me, ‘Of course they haven’t changed. We’re still hearing these stories.’”

How could this be? The explanation is multifaceted, according to Lipska and Kasper. 

“There’s just still a lot of holes in one who is covered by various protections and then how they’re covered and how they access this,” Lipska said. “It’s not as simple as just showing up at the pharmacy and asking for your insulin, and the sticker price is there, and that’s what you pay. It’s extremely complicated.”

For starters, millions of Americans remain uninsured and have to pay the cash price for drugs. While drugmakers make flashy announcements about helping those people, the process of enrolling for savings cards to get those lower prices can be a nightmare; one congressional investigation found in 2023 that uninsured patients were still paying an average of nearly $100 for insulin despite Eli Lilly’s pledge that its generic product would be available for $25 per vial. Even if you manage to get your discount card, the same investigation found that Lilly introduced several restrictions based on geography and needs, and that the specific insulins that are subject to these voluntary price caps are often out of stock at the pharmacy when people go to fill a prescription (while the more expensive brand name version is consistently available).

And then there are the loopholes in those existing federal and state laws. The federal $35 cap applies only to Medicare; the private insurance that covers half the country is excluded. (There have been bills proposed to extend the cap to private coverage, but they have not gone very far in Congress.) The state laws that do exist typically apply only to small-group and individual plans that are subject to state regulations; larger self-funded plans are exempt from state mandates and currently have no enforced limit on out-of-pocket insulin costs. One national estimate found that around 775,000 people who use insulin and had private insurance were protected in 2025 by the caps in states where they’d become law. But more than 1 million people in those states were not, nor were the 2.1 million commercially insured insulin users in states without a cap.

Even for people who benefit from the mandatory out-of-pocket caps instituted by Medicare and state laws, some of them still struggle to come up with $35 a month, given all of the other price hikes Americans have faced since the pandemic. They may have to navigate a complex process in order to actually receive the promised discount. And while a $35-per-month cap sounds straightforward, in practice, people can still face higher costs. Each individual person requires a unique amount of insulin and, for some of them, they might need two different kinds; the pharmacy may not have the right version and will make substitutions that lead to higher costs. That’s how one of Lipska’s patients ended up paying $140 per month for his insulin — despite being on Medicare.

“These things really add up…and frustrate people,” she said, “because what seems like a great idea is not actually how things work on the ground.”

The consequence is that, despite all the legislative activity and self-congratulations, millions of Americans are still at risk of serious health problems because they can’t get insulin.

The worst-case scenario is what happened to Alec Smith: Somebody is forced to spread their insulin so thin that they put their life is in danger. When diabetics have their blood sugar spike without insulin on hand, they can go into diabetic ketoacidosis, “which literally means your blood is turning into acid,” Kasper said. “I can speak from experience: it feels like you’re dying.”

To recover from an episode, they may have to go to the emergency room and end up taking even more insulin while eating in order to get their body back to equilibrium. But even short of an emergency, Kasper said, people who resort to rationing insulin can experience significant side effects. Prolonged high blood sugar due to rationing could lead to serious vision and nerve problems that were supposed to be a thing of the past once insulin became available.

Those were the scenarios that lawmakers wanted to prevent as they tried to make insulin more affordable. But they remain a very real threat for too many people.

The big idea that could make insulin truly affordable

If there’s good news to be found, it’s that the advocacy community is not giving up. They are sounding the alarm about the ongoing affordability crisis and urging Congress to act. But they do worry that the issue has become less urgent: The truly comprehensive mandate for insulin out-of-pocket caps, which is typically introduced every Congress, was delayed this time. “This year, neither bill was introduced for over a year, and it was very concerning,” Kasper said.

But they are also placing bets on something much more profound than a price cap: the government itself procuring or even producing insulin and making it available to people at the lowest possible cost. 

“This idea that medicines should be for people and not for profit I feel like is really catching,” Kasper said. 

Some states are taking baby steps in that direction: Oklahoma passed a bill this year that would offer government funding for private companies that pledge to produce insulin and sell it at a low price. 

Others are going bigger. As I reported in 2023, California has launched a program that is eventually intended to lead to government-owned facilities producing insulin to be sold on the cheap, though it will take years until publicly owned production begins. The New York state Senate approved a similar plan this year, although funding for it didn’t make it into the state’s budget. Kasper said that she is optimistic the project will make it into next year’s spending plan, and says a lot of work is currently being done to ensure it happens.

In the future, public manufacturing could also help to address the supply problems that continue to pop up. There are only three insulin manufacturers in the US right now. As those for-profit pharmaceutical companies place more emphasis on GLP-1 production — which often takes place in the same facilities as insulin production — nonprofit initiatives could fill in the gaps.

“In this time of tariffs even, with these concerns around shortages and rising prices and the corporations having this monopoly and price gouging ordinary Americans for decades now, as we’ve seen on insulin, let’s just make it ourselves,” Kasper said.

This is a long-term vision; if you need help right now, T1 International has resources to help people get health insurance, access discount programs, and even find emergency insulin. But a more affordable future for insulin will only be possible if lawmakers feel compelled to act. Public outrage created one round of reform, but the risk of political complacency is real. So if you care about this issue and want to get involved, T1 International also has a tool kit for people who want to advocate for policy changes, including around public pharma.

“The general communication that we’ve been hearing is that the insulin price crisis is over,” Kasper said. “Everyone should have access to a $35 copay for insurance. As the data shows, as we’ve been talking about, that is just not true.”

If you felt a gust of wind late on Friday, it came from hundreds of immigration lawyers who simultaneously exhaled after the Supreme Court announced it would not hear a case known as Genalo v. Black. That case asked whether certain immigrants could be held in immigration detention facilities indefinitely without ever receiving a bond hearing or other opportunity to seek temporary release. And if the Court had decided this case, its recent immigration decisions suggest that most of the justices may have blessed indefinite detention.

The Black case originally concerned two lawful permanent residents of the United States who were arrested during the first Trump administration because of crimes the two men committed years before, and that they had already served prison sentences for committing. 

These immigrants were arrested because of a federal law, which provides that the federal government “shall take into custody” noncitizens who committed certain crimes or who are accused of involvement with terrorism so that the government may begin deportation proceedings against these individuals. This law does not allow these noncitizens to challenge their detention while their deportation case is pending. (Notably, the relevant federal law sometimes allows an immigrant to be detained without hearing if they are merely suspected of terrorism, which means the person could be detained indefinitely based on false allegations.)

Both men sued, claiming that it is unconstitutional to hold them for too long without some kind of hearing to determine whether they should be released after paying a bond. And a federal appeals court agreed, holding that “due process bars the Executive from detaining such individuals for an unreasonably prolonged period…without a bond hearing.” 

Thus, Black was more than just another flash point in America’s debates over immigration. It was a test of one of the most fundamental protections that a liberal society provides to its residents — the right not to be locked up, forever, without any way to challenge the basis of that detention. 

Now that the Supreme Court has removed the Black case from its own docket, this appeals court decision will not be formally repudiated by the Supreme Court. So immigrants who are detained indefinitely may still ask courts to give them a bond hearing — at least for now.

Still, the Court’s nondecision in Black leaves a cloud of uncertainty hanging over similarly situated immigrants. The Trump administration will no doubt continue to detain people for prolonged periods of time (most of whom won’t be able to obtain a gubernatorial pardon, like one of the men was in this case). And the question of whether these immigrants can be detained forever will remain unresolved until the Supreme Court takes up another case involving a similarly situated immigrant.

Why the case was dismissed 

Both men at the heart of the Black case spent significant amounts of time in immigration detention facilities years after serving sentences for crimes. One, a Jamaican citizen named Carol Williams Black, spent seven months in an immigration detention facility. The other, Keisy Guerrero Mariano, spent 21 months in detention without a bond hearing or other procedure that could allow him to seek release while his case was pending. These 21 months in immigration detention more than doubled the time he spent in New York state prison for his actual crime — he pled guilty to second-degree assault after he got into a fight outside a New York City restaurant.

But there is no longer a legal dispute about what should happen to either man — in the words of the law, both cases are now moot. Black has since been released from detention and says that he “made the difficult decision to leave the United States” and has no intention to return. And the Trump administration’s case against Guerrero Mariano went up in smoke in late August, after New York Gov. Kathy Hochul gave him a pardon — thus wiping away the conviction that allowed the Trump administration to detain Guerrero Mariano in the first place.

Guerrero Mariano is one of six individuals that Hochul pardoned at the time. In her official announcement, she said that she did so because these six people “have demonstrated remorse, exemplified rehabilitation and displayed a commitment to improving themselves and their communities.”

There’s also a great deal of evidence indicating that Hochul pardoned Guerrero Mariano in order to resolve his immigration case — and potentially keep the Supreme Court from deciding Black. Among other things, the actual pardon document signed by Hochul specifically states that it will grant Guerrero Mariano “relief from removal.” Guerrero Mariano’s lawyers had sought this pardon, and Hochul granted it mere days before his brief was due in the Supreme Court. 

That brief’s very first sentence, moreover, states that Hochul “granted Respondent Keisy G.M. a full and unconditional pardon of the state criminal conviction that is the sole basis for his removal,” and it urges the justices to dismiss the case as moot.

Last Thursday, Trump’s Justice Department joined a stipulation calling for the justices to dismiss the case against Guerrero Mariano, most likely because even they recognized that a lawful permanent resident with no conviction may not be detained. The Court granted that request on Friday.

So that’s good news for Guerrero Mariano, who presumably will now be able to live freely in the United States. 

Guerrero Mariano’s lawyers were right to fear a trip to the Supreme Court

It’s possible that the Supreme Court might have agreed with the lower court’s holding that immigrants who are detained for an “unreasonably prolonged period” have a constitutional right to a bond hearing — this Court has, on rare occasions, broken with the Trump administration in immigration cases. But those decisions are few and far between.

Among other things, this Court has allowed President Donald Trump to deport immigrants back to war zones. The Court’s Republican majority effectively nullified the Convention Against Torture, a treaty that is supposed to protect immigrants from being deported to nations where they will be tortured. The Court also blessed so-called Kavanaugh stops, which permit suspected immigrants to be racially profiled so long as their race or ethnicity isn’t the sole reason why they are stopped by law enforcement.

Moreover, while the US Court of Appeals for the Second Circuit’s decision in Black — the one holding that immigrants like Guerrero Mariano may receive a bond hearing if their detention is “unreasonably prolonged” — reached a plausible legal conclusion, this conclusion was not a slam dunk under the Supreme Court’s previous precedents. And that means that the justices could have easily come up with a reason to reverse that decision if they wanted.

On the one hand, the Court said in Zadvydas v. Davis (2001) that “a statute permitting indefinite detention of an alien would raise a serious constitutional problem.” On the other hand, the Court held in Demore v. Kim (2003) that an immigrant similarly situated to Guerrero Mariano may “be detained for the brief period necessary for their removal proceedings.” 

Read together, these decisions suggest that the government does have some power to detain people while their immigration case is pending. While Zadvydas indicates that this detention can go on for too long, the Court has never established when an immigrant’s detention has stretched too long. So that means that the justices would have a fair amount of leeway to allow prolonged detentions without defying any of their previous precedents.

So it’s clear why everyone from Guerrero Mariano’s lawyers to Hochul seemed concerned about what would happen if the Court decided this case. In the worst-case scenario for immigrants, Trump could have gained the power to lock many people up for years without any meaningful way to challenge their detention. In the best case, the Supreme Court may have reached the same conclusion as the Second Circuit — but even that decision was vague about how long an immigrant must be detained before they must receive a bond hearing.

In any event, the bottom line is that the Court’s decision to dismiss the Black case is good news for Guerrero Mariano. But that dismissal merely delays a reckoning for other immigrants who face prolonged detention while their case is pending.