It’s an odd definition of “victory.”

On Wednesday, Instagram head Adam Mosseri was supposed to be back on the witness stand in federal court in Oakland, California, where 29 states were trying to prove that Meta designed Facebook and Instagram to hook young people and misled the public about the risks.

He had testified Tuesday defending the company’s safety record and pushing back as a plaintiff’s attorney kept pointing out how few teenagers actually used Meta’s tool designed to encourage kids to take breaks from the app. Mosseri was scheduled to return and finish.

He never did.

By Wednesday morning, Meta had reached a settlement worth up to $17.1 billion, and a trial that began the week before simply stopped.

California Attorney General Rob Bonta called the settlement a watershed, and said it was “telling” that Meta settled midway through Mosseri’s testimony and before Meta CEO Mark Zuckerberg could be called. Meta’s lawyers, he said, “were dying to have [Mosseri] get off the stand.”

That may be true. But now no one will hear the testimony that Bonta says Meta was afraid the public would hear.

Zuckerberg was expected to testify within days. Thousands of internal documents had been identified as potential exhibits. Now that testimony and evidence that hasn’t been presented may never see the light of day.

What Meta Will Pay and Change

The settlement landed on Chief Judge Tvonne Gonzalez Rogers’s desk at 6 AM Wednesday. She took the bench later that morning and questioned the lawyers for about an hour. She said she hoped to approve the settlement. Until she does that, the settlement isn’t a done deal.

The agreement reaches beyond the states involved in the trial. Forty-seven states, the District of Columbia and several territories have signed on. Texas negotiated its own $1 billion deal with Meta.

Two states stayed out of this settlement entirely. As we’ve reported in recent episodes of Question Everything, New Mexico just won its own nearly $1 billion case against Meta.

New Mexico Attorney General Raúl Torrez said the new settlement missed protections his case had won, including safeguards against adults targeting children. But he added the changes represent “real progress” to protecting kids online.

Florida begged off for this round, calling the multi-state payment “peanuts” and pledging: “We’ll see them at trial.”

Meta will pay a guaranteed $12 billion over ten years, with more possible if other states join and competitors accept similar terms.

Meta also agreed to change how its apps work for teens in the US. The company will interrupt endless scrolling, limit teen use to two hours a day, shut the apps down overnight and silence notifications during school. It will hide like counts, offer a feed that isn’t personalized by an algorithm and ban cosmetic surgery filters. It will also strengthen age verification, connect a teenager’s multiple accounts and require verified parental approval before certain protections can be switched off.

If you have a 14-year-old, their screen time is about to change dramatically.

But the settlement brought an end to the trial before the public could learn what else Meta might be hiding.

Trial Sets No Precedent

There will be no verdict. No findings of fact. No ruling on whether Meta misled the public about what it knew. Again, Zuckerberg and many other executives never took the stand. Only a fraction of the thousands of documents on the exhibit lists made it into open court.

Some could surface in other lawsuits. (Meta still faces thousands of personal injury lawsuits and cases from school districts.) But this trial was supposed to do something a document leak can’t. It would assemble the evidence, put witnesses under oath and ask a jury to decide what Meta knew and whether it told the truth.

The settlement also leaves a larger legal fight unresolved. Meta has argued that Section 230 and the First Amendment shield it from cases like this one. Bonta said the deal showed those defenses aren’t cast in steel. But he acknowledged that a settlement carries no precedential value. So the next social media company to be sued can still raise the same defense. It would be like starting all over again.

Judge Rogers, for her part, sounded relieved that she wouldn’t have to finish the trial and write it all up. But she did raise a concern about Meta’s use of attorney-client privilege.

During the last 48 hours of the trial, plaintiffs’ attorneys presented an Instagram strategy deck from September 2023. The deck was stamped attorney-client privileged, with several items blacked out, even though it looked like a product review, not a request for legal advice.

On a page listing Instagram’s goals, a safety target was marked “at risk” because “Take a Break defaults” were being blocked, meaning that folks inside the company wouldn’t approve it at that time.

Take a Break is the feature Mosseri was being grilled on that nudges teenagers to put down their phones. Mosseri said that they made the feature a default in 2024, but the exhibit shows his team wanted to do it in 2023, and the change was being blocked inside Meta.

Another deck, also marked privileged, laid out where Instagram might “lead” on teen protection. But an email from July 11th, 2023 showed that their definition of “lead” meant that they were able to convince people they were improving rather than actually doing better than competitors.

Those documents reached the jury just hours before Meta settled.

We don’t know what was in the documents the jury never saw.

Meta says its employees are loose with how they talk to each other. (Think about how you communicate with colleagues at work.) It says the states cherry-picked phrases to make them look bad and that its teen-safety work leads the industry.

Bonta had a not-so-loose phrase for the protections that Meta rolled out after the states sued. He called them “tissue paper defenses.” He said a child could override those safeguards with a swipe of a finger. Under the settlement, that changes. Parents will have to be verified as parents before a default can be lifted.

He also described the agreement as “a floor,” not a ceiling. State legislatures are still free to go further, and his office is still suing TikTok.

Settlements are bargains. The states secured billions of dollars and protections that could change how millions of teenagers use social media. Meta avoided the risk of a verdict and years of appeals.

And the public may never know what other secrets the states were waiting to expose in court.

Kevin

Subscribe now

Ava experienced mental health issues due to social media use as a teen herself, and is a member of the cohort Haidt’s work is most aimed at. His bestselling book is called The Anxious Generation, and yet I’ve never heard Haidt interviewed by someone from that generation. Ava is a policy advocate for social media safety in Washington — in many ways you’d think she and Haidt are aligned, and they are — but she also has bones to pick with him. She thinks his proposals focus too exclusively on young people. What about all the boomers who are having their minds addled on Facebook?

Ava questioned Haidt about this live on stage in New York this spring, in the wake of the two huge jury verdicts against Meta and YouTube, and I found it to be a fascinating, exciting, memorable interview — and one where I got to see aspects of Jonathan Haidt I hadn’t before — so we’re sharing it on Question Everything this week.

I hope you like it. And if you do, please check out the rest of Left To Their Own Devices. It’s really good.

Brian

Subscribe now

The headline printed across most major news outlets this week is that Meta is being ordered to pay $567 million to New Mexico. That’s on top of the $375 million in penalties that a jury already decided Meta has to pay for violating state law in the first phase of this trial, for a total of nearly a billion dollars. It’s the largest amount of money Meta has been ordered to pay as the result of a trial, ever.

But the most interesting part of this ruling isn’t the money. For a company that made over $200 billion in revenue just last year, this hardly seems like a big financial loss.

To me, what’s more novel about the judge’s decision are the mandated changes — or injunctive relief — he’s requiring Meta to make to Facebook and Instagram in New Mexico. This is the first time a judge has ordered a social media company to make these kinds of structural changes to its platforms to protect kids.

Tech companies like Meta are typically able to avoid liability by invoking Section 230. Listeners of our show are familiar with this 1996 federal law that protects internet platforms from being held legally responsible for content posted by third-party users. The tech companies can claim that it’s the content, not their platforms, that’s causing the damage. But the state argued that Meta has caused a “public nuisance” to the people of New Mexico (that’s the legal term) and therefore should be forced to change to reduce the harm to people in their state.

The judge agreed. This case got around Section 230 because New Mexico argued that it was the way Meta designed its products that caused harm, not the content. The state was allowed to go to court on those grounds, and a decision was issued. The judge even borrowed an analogy that came up in trial and compared Meta to a factory, writing:

“Just as noxious pollution produced by the factory can harm the common public right to reasonably clean air, the harmful effects of Meta’s platforms on children do not stay contained by its platforms and, instead, migrate to the internet as a whole and, perhaps most concerning, to the real world and create a common, societal burden on and harm to the affected children and their families and schools, as well as hospitals and law enforcement.”

I want to break down some of the key changes the judge is mandating Meta make to Facebook and Instagram in New Mexico.

The first bucket of changes is connected to features that the state argued were designed to addict teens to these apps.

The state wanted Meta to remove infinite scroll, autoplay, and like counts for teens, and also restrict when Meta can send push notifications to teens, and how long they can spend on the apps. The state also asked that Meta be forced to change its algorithm to prioritize “integrity over engagement.”

Judge Biedscheid granted half of these requests by the state, and denied the others.

The judge denied the state’s requests to remove autoplay and infinite scroll and is not forcing Meta to make changes to its algorithm. He wrote that, “autoplay, infinite scroll, and algorithmic content are features with clear Section 230 and First Amendment implications.”

Having videos begin to play as soon as a user opens the app, allowing users to scroll from video to video without stopping, and the design of the recommendation algorithm that pushes content to users… the judge said that all of these are effectively editorial decisions the company is making about displaying content from its users. That’s what Section 230 is about, and therefore, Judge Biedscheid said, those elements are protected, and he can’t force Meta to change them.

However, the judge did order Meta to remove like counts, restrict push notifications, and he told Meta to put a time limit on teens’ use of Meta’s apps. If this decision holds up (Meta has vowed to appeal ), teens in New Mexico will no longer be able to see the number of likes on a post on Facebook and Instagram by default; they will not receive push notifications during school and sleep hours; and their access to the apps will be capped at 90 hours per month.

The judge said these features are proven to be addictive, and they do not have to do with user content on the platforms, which is why he is legally allowed to make this order. I’ve never heard of mandates like these that restrict kids’ access to major social media platforms in the United States. If this order is implemented, this does seem like a big change.

Judge Says a Law from the 1990s Interferes with Age Verification

A big topic in last week’s episode was whether Meta would be required to improve its age verification. It’s Meta’s policy that you have to be 13 to use their apps, but the state argued Meta isn’t effective enough at knowing who is actually under 13. They asked the judge to order Meta to get to 99% accuracy at knowing their users are at least 13 years old.

The judge didn’t go for it. He is requiring Meta to improve its models for detecting when kids are under 13 on their apps over the next two years, but doesn’t provide specific metrics for how to do that.

The reason the judge says he’s barred from ordering more strict age verification is another 1990s-era federal internet law: the Children’s Online Privacy Protection Act, or COPPA. It’s a law that prevents internet companies from tracking or maintaining any data about kids under 13. Meta argued in court that COPPA has prevented them from doing the kind of age checks the state was asking for. Meta’s lawyer Alex Parkinson said this about COPPA in his opening statement:

“The state wants a system that verifies that users are at least 13 years old to 99% accuracy. Gee, what a great idea. I wish we had thought of that. We did. We tried for four years to build that kind of classifier, and Your Honor heard the technological problem is COPPA, which the state has no answer for.”

The judge agreed with Meta on this one. In his decision, he wrote, “because of the COPPA Rule, it is the Court’s conclusion that it cannot order Meta to request children to submit personal data or be passively tracked online, even for age-verification purposes.”

COPPA creates a legal Catch-22 around checking kids’ ages. It was designed to protect kids, just as age checks are meant to protect kids, but COPPA actually bars internet companies from collecting the information they need to verify the ages. That’s because COPPA says that any information internet platforms collect on kids has to be destroyed.

The FTC, which enforces the law, released a statement earlier this year saying they won’t enforce COPPA if the company is tracking kids’ information for the purposes of age verification. But the law hasn’t been changed. The agency just said, in effect, that they promise to look the other way. It’s not very binding, and it wasn’t enough to give comfort to this judge that he could issue his ruling in a way that violates COPPA. So Meta has to get better at knowing kids’ ages by scanning its posts (like if a kid is wished a happy 12th birthday on Instagram, that’s a big hint they aren’t 13). But the judge has a pretty loose definition of what “better” means.

End-to-end Encryption Doesn’t Change

The state also asked the judge to order Meta to ban end-to-end encryption for teens under 18 in New Mexico. That’s the technology that makes it possible for only the sender and receiver of a message to be able to read it. Not even the company that hosts the messages can read them.

I spoke to one of the state’s witnesses, Dr. Brian Levine, a computer scientist at UMass Amherst. He co-directs something called the Rescue Lab, which designs tools used by law enforcement to protect kids from exploitation online. He told me that end-to-end encryption creates a place where kids can talk to predators in secret. He called it a “black box of abuse.” Dr. Levine told me that banning end-to-end encryption is the most important thing on the state’s list to protect kids.

The judge did not do it in full

He allowed Meta to keep end-to-end encryption on Facebook Messenger and WhatsApp . He gave two reasons. First, he didn’t see enough evidence that many New Mexico teens were using Facebook. They are way more active on Instagram. Second, he heard testimony about the privacy benefits of the technology, and he felt that everyone in New Mexico, even teens, has a right to that privacy that comes with end-to-end encryption.

But here’s the twist that came up during the trial. Meta disabled end-to-end encryption on Instagram messages in May, while this trial was happening. The company said not enough people were using the feature. Though Meta’s timing could be seen as an attempt to appease the court. The judge ordered that Meta needs to keep end-to-end encryption shut off on Instagram for teens for at least the next five years.

There’s a lot more in the judge’s 68-page ruling, which we discuss in this week’s episode.

You can also read the ruling for yourself here. If you do, let us know what parts you found most interesting or surprising. Do you think this will keep kids safer? Shoot us an email here or reach out on Instagram @questioneverything.podcast to tell us your thoughts.

Zach

Subscribe now

Ellison used the column to argue that he is not trying to politicize the news organizations that are coming under his control — CBS News and potentially CNN. The format allowed him to do that without any pesky follow-up questions, challenges to his assertions, or alternate points of view. He was in control.

Subscribe now

Meanwhile, in news story after news story in the Times, Ellison has refused to talk to the paper.

➡️ In July 2025, when the paper profiled him as he took control of Paramount: “Mr. Ellison, who declined to comment for this article…”

➡️ In September 2025, when the paper covered his Hollywood spending spree: “Paramount Skydance and Mr. Ellison declined to comment for this article.”

➡️ In March 2026, when Secretary of Defense Pete Hegseth said, “the sooner David Ellison takes over [CNN], the better.” The paper noted: “A representative for Mr. Ellison declined to comment.”

The only Times interview I can find that David Ellison said yes to since Trump’s second term began allowed him to celebrate his purchase of U.F.C. rights:

The Ellisons’ company is currently being sued by 12 states and the Writers Guild for violating antitrust laws with its proposed Warner Bros. Discovery merger, and every week brings new evidence of the carnage his leadership has wrought on CBS News and 60 Minutes. And this is the moment that the Times is letting Ellison write a first-person retort in its opinion section?

David Ellison is one of the main people responsible for the fast-moving media capture that is taking hold in the U.S. I’ve interviewed experts in media capture from other countries who are “flabbergasted,” as one of them told me, at how quickly it’s happening here. We covered the very cozy relationship between David Ellison, his father Larry, and Donald Trump in our recent episode about Larry and Donald’s bromance.

Why is the paper of record — whose publisher admirably defends independent journalism — giving this man free rein to sugarcoat what he’s doing? It undermines independent journalism at the Times and in the U.S. media writ large.

I know the opinion section aims to publish a diversity of viewpoints, yadda yadda yadda. But this isn’t a diverse viewpoint. It’s the subject of critical Times reporting being allowed to do an end-run around that reporting. It reminds me of all the celebrity and athlete “documentaries” that hand over editorial control to the celebs.

I want to hear from Ellison, but in an interview with an independent journalist, not a crafted essay with no pushback. Put him in front of Lulu Garcia-Navarro! But he won’t do that if he’s ceded a bunch of inches in the paper to say what his heart desires.

This whole situation undermines the very claim Ellison’s trying to make in his column: he doesn’t respect independent journalists if he can’t even face an interview with one. Come on, man — that’s what the money is for!

Brian

Trump and Ellison are playing a huge part in the current consolidation and capture of American media. After Emily got a couple of insider tips that the two men actually like each other — as opposed to their relationship being one solely based on political or monetary convenience — Emily and her co-reporters, Jessica Toonkel and Annie Linskey, dug into the nature of their relationship.

What they uncovered helps shed light on how Ellison’s relationship with Trump appears to be different — deeper — than the president’s relationship with even other tech billionaires, like Elon Musk.

Some of the revelations the Journal reported include:

• A secret $45 million payment from Ellison to a nonprofit political group supporting Trump.

• A time when Ellison showed up for an event at the White House without photo ID, and President Trump personally intervened with security to get him through.

• Visits by Ellison to Mar-a-Lago and lots of phone chats about how Israel is great and the media sucks.

• An ethically questionable phone call from Larry Ellison’s son, David, to a Paramount board member.

• A promise by Ellison that Paramount would overhaul CNN if it’s able to gain control of Warner Brothers Discovery. Paramount denies it.

Emily broke down how she and her colleagues chased this story down, what surprised her most in the reporting, and why a friendship between a media mogul and a president is never just a friendship.

And the scoops keep coming for Emily. Today, she and her colleagues have an exclusive story on President Trump’s fundraising operation, and a woman nicknamed the “princess of darkness,” who helps extract hundreds of millions of dollars from companies for Trump’s pet projects.

Brian

P.S. To see more of our videos, check out Question Everything on Instagram. (@questioneverything.podcast)

Like usual, I learned something new when putting together this week’s episode of Question Everything: “The Sunscreen Paradox.”

This is a known scientific phenomenon that has shown up in multiple dermatological studies: people who say they use sunscreen regularly tend to have higher instances of skin cancer.

Counterintuitive, right? And also potentially alarming. We grow up being told to put on sunscreen in the sun. Have we been tricked into giving ourselves cancer?

That’s what former Florida Congressman Matt Gaetz asked a few weeks back on his One American News Network show.

“Have I been lied to as a Florida man about sunscreen protecting me from cancer?” he asked epidemiologist Nicolas Hulscher.

“Yes, it appears we’ve all been lied to, unfortunately,” Hulscher responded.

But no: we have not been lied to. Enter…The Sunscreen Paradox.

The reason people who report using sunscreen regularly also report higher rates of skin cancer is because they likely go in the sun more overall, and they tend not to use the sunscreen properly. You’re supposed to use at least an ounce on all uncovered skin and reapply every two hours, or after sweating or swimming. People who say they use sunscreen regularly may put it on when they get to the beach, but then maybe they don’t reapply it. They have a false sense of protection and stay out in the sun longer – and increase their likelihood of skin cancer.

When researchers follow people who use sunscreen properly, The Sunscreen Paradox goes away.

In this week’s episode, we explain how Nicolas Hulscher has been using this paradoxical scientific phenomenon to mislead people into thinking that sunscreen can give you cancer. It’s a deception that’s gotten so much traction in recent weeks that NewsGuard, the fact-checking company, recently named it their False Claim of the Week. And he’s spreading this lie on the platform I’m using to send you this newsletter: Substack.

Hulscher collaborates with an anti-vaccine cardiologist named Dr. Peter McCullough on a Substack newsletter called Focal Points that NewsGuard has identified as the source of many false health claims. And this one about sunscreen was among them.

If I’m being honest, I don’t think about sunscreen all that much (though I have tried to get better about applying it daily!) But the story has raised some interesting questions I never considered about the safety of sunscreen. Not that it causes skin cancer, but questions about whether it might cause other health-related issues. Right now, people are commenting on our Instagram account to raise concerns about sunscreen and say we should be careful about dismissing them.

“I don’t know why you say this is Maha conspiracy. I do believe sunscreen can be harmful,” wrote thehunnybe. “The chemicals in sunscreen are absorbed into your body. Your skin is the largest organ. Sunscreen confuses your body. It’s not viral studies that have made me feel this way. It just makes sense to me.”

Also a friend of mine, who’s a regular listener, texted me to say:

My friend told me he and his family have used only zinc-based sunscreens, which sit atop the skin rather than allowing chemicals like oxybenzone to be absorbed within the skin.

So I reached back out to the health editor at NewsGuard I interviewed for this episode, John Gregory, who’s spent a decade or so tracking and debunking medical misinformation online. What did he know about other, non-cancer concerns about the effects of sunscreen?

I’m just going to share his whole response, with his sources, here:

Health and medical authorities have consistently stated there is no evidence that chemical sunscreens actually cause hormone disruption. The concerns are based on animal and lab studies testing concentrations of ingredients higher than what anyone is exposed to with sunscreen.

https://www.cancer.org.au/iheard/do-sunscreens-contain-endocrine-disruptors

https://med.stanford.edu/news/insights/2025/06/sunscreen-science.html

Key section here: "Organizations that have raised concerns about oxybenzone typically cite studies done in rats, where the rats were actually fed oxybenzone. It would take an individual 277 years of sunscreen use to achieve the equivalent systemic dose that produced effects in these rat studies, according to a 2017 study in the Journal of the American Academy of Dermatology."

https://www.health.harvard.edu/healthy-aging-and-longevity/the-science-of-sunscreen

I’m not aware of any consensus opinion about whether zinc-based sunscreens are more or less effective than chemical sunscreens. There was an Oregon State study in 2021 which said zinc-based sunscreens appear to lose effectiveness more quickly: https://news.oregonstate.edu/news/osu-study-after-two-hours-sunscreen-includes-zinc-oxide-loses-effectiveness-becomes-toxic

If zinc-based sunscreens that sit on top of the skin need more frequent reapplication, that could affect a user’s long-term protection against skin cancer.

277 years!!!

Meanwhile, numerous other studies have shown that proper sunscreen use can decrease your risk of various skin cancers. (See: Sunscreen Use and Subsequent Melanoma Risk: A Population-Based Cohort Study; The efficacy and safety of sunscreen use for the prevention of skin cancer.)

Since so much of this rests on sunscreen being used properly, I’m going to share guidance for proper application that I got directly from the American Academy of Dermatology:

For adults, use a broad-spectrum formula, at least SPF 30, on all exposed skin 15 minutes before you head into the sun. Most adults need to use at least one ounce – a shot glass – to all skin not covered by clothing. Also, make sure to reapply every two hours or after swimming or sweating. For more info, check this out: Sunscreen FAQS from the American Academy of Dermatology

I’m curious: have you seen these claims about sunscreen circulating on your social media feeds, or in the comments? Now that I’m paying attention, I’m hearing more and more about it. Let me know if you’re seeing it come up, and what you think.

Brian

Subscribe now

The Internet Crimes Against Children investigators, like Jonathan Tally and Rickey Kitch in this story, are part of a national task force program. There are 61 coordinated ICAC task forces working across all 50 states. Jon and Rickey operate in New Mexico. Reporting this story, I got to thinking about the job Jon and Rickey do as ICAC investigators, how appreciative I am for their work, and how it’s a job for a select group of thick-skinned people.

I talked to Jon for less than an hour, sitting in ICAC’s dedicated darkroom for reviewing child sex abuse material, called the “windowless room” in the story, with producer Zach St. Louis. Listening to him, I came away with a much clearer picture – not of the toughest guy on the planet (no shade, he’s definitely tough), but of the nuanced sensitivity that is being wrestled to the ground each day, to perform the kind of work Jon has dedicated his career to.

As the head of ICAC, Jon told me some of the lessons that he tells new investigators in training. These are things he learned firsthand.

Rule One: Don’t Bring Things You Like into the Windowless Room

Jon stays caffeinated by drinking Monster Energy drinks. But conventional wisdom for ICAC investigators says to avoid drinking anything that you habitually like while reviewing images and materials in the room. In fact, the advice is to quite literally keep everything off the desk that you normally would at a desk job. No phones, no key chains, no Monster drinks — nothing that you might glimpse or associate with seeing traumatizing images of kids.

It’s not a casual thing. Jon told me that looking at child sex abuse material can be so triggering as to jolt you back to the image, in spite of yourself, by just having things you like nearby doing the work.

Lots of people say you should separate work and life. This is the best example I’ve ever heard.

Rule Two: Mute Button

Back when Jon was a newbie at ICAC, he told me he’d only been on the job for maybe two weeks when he got his first tip from the National Center for Missing and Exploited Children. This was during COVID, and he says he was one of the only agents in the office at the time.

The tip was a video, which I didn’t have him describe in detail, but for context, it involved a baby being sexually assaulted, which Jon only learned by playing it. He watched and listened, which is a mistake.

After a few seconds, he stopped, panicked, thinking: How will I be able to get through watching stuff like this?

Desperate for advice, he went to his commander asking, “Tell me a secret.”

His commander told him, “Watch with the sound off.” Which Jon has done every time since.

Rule Three: Review Materials in Small Doses

Jon says, only work in the windowless room for an hour at a time, tops.

When Jon testified in the trial against Meta, he talked about the mental health struggles he’s faced as an ICAC agent and how common they are. Though not all ICAC investigators are former police officers, Jon is, and he noted on the stand that the burnout rate of police officers pales in comparison to the burnout rate of ICAC investigators, which he puts at around two years, in his experience.

Watch Jon on the stand discussing some of the difficulties of the job:

Which brings us to Rule Four: Use the Hotline

There’s a dedicated mental health provider available 24/7 for these investigators. Jon’s on a first-name basis with theirs, Carol. He says his previous commander would regularly encourage people to use it, to not try to muscle through the job alone. Jon told me he does use her, and has probably called the hotline himself, maybe eight times over the past few years.

Finally, I wondered how people in this line of work who are parents handle having teenagers themselves. Do their kids even have phones?

Attorney General Raúl Torrez has two kids, a 16-year-old daughter and a 13-year-old son. His daughter had a Bark phone, which he could do a commercial for. It’s so up his alley. Bark phones are smartphones that are packed with parental controls that let parents monitor content, get 24/7 alerts, and change settings remotely.

His daughter just recently graduated to a more standard phone. His son, 13, has a watch for calling and tracking. No phone. The kids hate the stigma of not having phones like their friends, but the AG is immovable: “They had the misfortune of having a career internet crimes and child abuse prosecutor as a parent, so there’s not a lot of room for negotiation.”

Jon Tally has a 16-year-old daughter and 10-year-old son.

His daughter is about to turn 17, so she has a phone, and she has more freedom, though he still checks her phone now and then.

His son also has a phone — under tight surveillance by Jon. He checks his son’s phone a lot. “He knows he can’t be trusted,” Jon told me. He also knows his dad knows all of the tricks — how to find deleted messages, hidden apps. And at 10, no social media. His son can play Roblox, but not without Jon keeping watch. He tells his son, “ Dude, I have to monitor your stuff because there are bad actors everywhere.”

Robyn

Subscribe now

We are running the interview anyway. You can watch the full thing, unedited, and decide for yourself. Am I slimy?

I had initially reached out to Ashley because she is in a legal battle with Musk’s company, xAI, over sexually explicit deepfakes that its chatbot Grok made of her. The bot generated millions of these kinds of images of her and other women, and I wanted to hear how she planned to take on one of the most powerful companies in the world, run by the richest man in the world, who also happens to be her ex. I wanted to know how she planned to address one of xAI’s first volleys — invoking Section 230, the law that has protected internet service providers like xAI for 30 years from facing liability for horrible content on their sites. Ashley’s case is one of the first where a company is trying to use Section 230 to protect itself from being held liable for making a generative AI product that harmed people.

We talked about that, but I also asked about her MAGA past, because… journalism. I asked her about some of her past videos and tweets about migrants, trans people, lies about Haitian immigrants eating domestic pets, and the 2020 election. It was those questions that seemed to bother her most.

She said she wasn’t expecting the interview to go there, and demanded we shelve it. This surprised me, because Ashley’s been on something of an apology tour, going on lefty podcasts to say she’s remorseful for the hateful rhetoric she spouted during her years on the right-wing. In the lead-up to the interview, she told me to prepare as if there was nothing she couldn’t talk about. “I am good at pivoting,” she texted me.

We told her we planned to run the interview. Then she emailed, claiming she never wanted to kill it. She said she only wanted us to delay publication until she talked to our lawyers. We had connected her with them, but she never ended up doing the call.

People online have been debating whether they can trust Ashley’s transformation from MAGA to “woke,” and I think this interview will help you decide.

But also — what do you think of my decision to ask Ashley about her past rhetoric? And our choice to run the interview despite her protests?

Watch for yourself, and let me know what you think.

If you’re not already a subscriber, this is a good time to become one. Hit the subscribe button to get more of these conversations delivered to your inbox. Our fact-checker reviewed this live, unedited conversation. We were not able to independently verify St. Clair’s remark that Musk’s devices were seized during French authorities’ search of X’s offices. Likewise with St. Clair’s recollection that she saw a Grok-generated image of a 4-year-old girl in a Bluey microprint bikini, and her claim that immigrants were getting on planes with boarding passes that didn’t have any name on them. We reached out to Musk and xAI for comment about St. Clair’s allegations, but they did not reply.

Subscribe now

After 16 years in power, Hungarian prime minister Viktor Orbán was recently voted out in a stunning landslide. Suddenly, democracy-loving Americans are looking to Hungary for inspiration, instead of as a cautionary tale. How did Hungarians get rid of their leader and usher in a new day for democracy?

Yes, there was the political strategy of the opposition candidate, Péter Magyar. Yes, there was the organization of various civil society groups. But again and again I’ve heard experts giving credit to journalists for this outcome. And not in some high-mindedly vague journalism-is-essential-to-democracy-kind-of-way. There are two specific journalists, who uncovered a specific, juicy story about Orbán and his allies that people have been saying truly helped bring the prime minister down.

András Pethő and Zsuzsanna Wirth work at Direkt36 — one of a few tiny, independent news outlets left operating in Hungary, after Orbán’s decade-long capture of the media. In the weeks leading up to the election, they got a tip about a plot to try and sabotage the opposition campaign involving a shadowy political operative, a young IT guy known as “Gundalf,” a boat full of files floating in the Danube, and a secret camera hidden in the hole of a belt.

Marty Baron, former executive editor of the Washington Post, noted that András and Zsuzsanna’s reporting “played an essential role in the election and in safeguarding what was left of democracy.”

Péter Magyar, the new prime minister, called their story the Hungarian Watergate.

Since Viktor Orbán’s defeat, I’ve wanted to know – what was this story that helped bring him down about exactly? And how did the journalists uncover it?

So I asked Sean Cole to report it out—to talk with András and Zsuzsanna—and this week he tells the whole hair-raising story on Question Everything. You can listen here:

It’s like a real-life spy thriller. But also, there’s something for Americans to learn here, for sure.

Brian

Subscribe now

“I would like to thank President Trump for entrusting me with the privilege of being nominated to the Southern District of Florida, a district where I spent much of my career as an attorney,” Judge Kuntz told the committee. The Senate is still considering Kuntz’s confirmation.

Not long after that, an organization called the Freedom of the Press Foundation filed a judicial ethics complaint against Judge Kuntz, with the aim of getting the judicial authority in Florida to strip Kuntz of his gavel and law license.

“Kick him off the bench,” Freedom of the Press Foundation’s chief of advocacy, Seth Stern, told me.

President Trump wants this Florida judge to get a huge promotion. A free press organization wants the same judge barred from practicing law. What gives?

Well here’s the deal: Judge Kuntz recently ruled on a defamation lawsuit that was filed by, of all people, President Trump against, of all places, the board of the Pulitzer Prizes. Yes, the President is suing the people who bestow the most prestigious award in journalism. And while we all know he is on a legal crusade against the press – having sued the New York Times, the Wall Street Journal, the BBC, CBS, the Des Moines Register, among others – this lawsuit against the Pulitzer Board is definitely the most peculiar. Trump’s suing them for giving an award to coverage that was critical of him.

The case is still ongoing. And that’s in large part because of Judge Kuntz, who, along with two other judges, ruled in Trump’s favor after the Pulitzer Board moved to have the case thrown out. They said Trump could proceed with the suit.

Kuntz handed Trump a big win. And not long after, Trump handed Kuntz a nomination.

And that nomination didn’t just fall out of the sky.

“The Pulitzer Board did not know while Judge Kuntz was considering their case, that he had been seeking a nomination to the federal bench,” says Stern. At the same time Judge Kuntz was considering Trump’s case against the Pulitzer Board, behind the scenes, he was making his interest in a lifetime judgeship known to Donald Trump, trying to get Florida’s Republican senator to put in a good word for him with the president.

Stern and his colleagues believe that’s a serious conflict of interest.

“Seems pretty obvious that he should recuse himself or disclose the conflict but he did neither,” Stern told me. “Obviously, Donald Trump has a very involved role in selecting the judges who he commonly calls ‘his judges.’ And word that a judge had ruled in his favor on one of his many frivolous lawsuits is likely to be helpful in securing a nomination from the president.”

Stern said even if there wasn’t an explicit quid pro quo, “it looks pretty fishy.”

Maybe even fishier: Kuntz wasn’t the only judge on that panel who was seeking a nomination from Trump while ruling on the Pulitzer lawsuit. Judge Ed Artau also decided in favor of Trump.

“He not only ruled in favor of Trump, but he issued a concurring opinion in the case that read like a MAGA manifesto,” Stern said. “It called for the repeal of New York Times vs. Sullivan. It talked about fake news. It hit all these talking points.”

I remember reading this concurrence when it was issued, and seeing the phrase “FAKE NEWS” written just like that, in all caps – in a judicial opinion!

“It read like something Trump would put on Truth Social,” said Stern.

Shortly after that ruling, Trump nominated Judge Artau to the federal bench. He was confirmed in September.

I’ve been following Trump’s cases against the press pretty closely. I know how bad they are for the press. But I hadn’t considered this noxious byproduct: When a president is suing somebody personally, it gives judges an easy opportunity to curry favor with him, by issuing decisions they know the President will like.

The Freedom of the Press Foundation has been trying to stop the potential graft, by filing these ethics complaints. But according to reporting on Capitol Hill, it looks like Judge Kuntz is on track for Senate confirmation regardless. The judge told senators that Florida’s ethics rules did not require him to disclose his attempts to get the appointment from President Trump.

Meanwhile, the Pulitzer Board continues to fight. Oddly, the board isn’t a formal organization: It’s a group of 18 individuals, including New Yorker editor David Remnick, the former editor of the Boston Globe, Nancy Barnes, poet Elizabeth Alexander, bestselling nonfiction writer Katherine Boo, and novelist Viet Than Nguyen. They’re now in discovery, and are trying to extract information from Trump that he has long sought to keep locked up, including his tax returns, financial information, and psychological and medical records. It’s possible, if the Pulitzer Board members are successful, that they could eke out sensitive information about Trump that reporters have long been trying to get leaked to them. A possible silver lining.

Overall, it’s a bizarre case that should have no legal standing. But it does raise some interesting ideas about how individual facts sometimes don’t add up to truth. That’s what we get into in this week’s episode, where I share an interview I did last year with Semafor’s Ben Smith, who’s been reporting on the Pulitzer case. Check it out.

Brian

Subscribe now

Like many of us, Gordon Pennycook, an associate professor of psychology at Cornell University, had the intuitive sense that people who believe conspiracy theories rarely change their minds about them. In fact, Gordon’s brother is a believer in QAnon-esque conspiracy theories, and Gordon has not been able to change his mind. But Gordon, bless him, wanted to test this assumption scientifically. So he and a team of researchers at Cornell devised an experiment.

They recruited thousands of people, and asked the participants: “From your perspective, what is a significant conspiracy theory that you find credible and compelling?” They also asked the participants to rate on a scale of 1 to 100, with 100 equaling total certainty, how much they believed their conspiracy theory.

The researchers then fed each participant’s theory, and the reasons they gave for believing it, into an AI model — ChatGPT 4 Turbo. They told the AI — which they branded Debunkbot — to use strong facts and evidence to counteract those views.

For example, one person believed 9/11 was an inside job by the government because World Trade Center Building 7 collapsed without even being hit by a plane. Debunkbot explained that the National Institute of Standards and Technology did extensive investigations which found that Building 7 fell because debris from the North Tower hit it, causing fires. The participant came back wondering why the hijackers were let through security so easily; the bot explained the timeline of airport security measures and how most of what we experience flying today came about after 9/11.

This was limited to just three back-and-forths — three inputs from the participant, and three responses from the chatbot. Once that was over, the participants were asked to rate their belief in the conspiracy again.

The results were stunning. Gordon says about a quarter of people who believed a conspiracy theory at the start of the conversation with the chatbot no longer believed it was true at the end of the conversation.

Just think about that: an 8-minute conversation with an AI model, and someone who believed that the U.S. government killed thousands of Americans by demolishing the World Trade Center and then covering it up, could come away realizing that probably wasn’t true.

Gordon and his team couldn’t believe the results either. “We were like, there’s a mistake. This is messed up,” Gordon told me. “And then we went through it again, of course. We replicated it. It’s the real thing.”

And there were other crazy findings:

• Two months after the conversation with the chatbot, the researchers followed up with the participants, and their belief in the conspiracy theory hadn’t come back. They truly did change their minds.

• When researchers asked the participants about other conspiracy theories — not the one the AI had specifically debunked — they believed those less, too. Their belief wasn’t as diminished as it was in the core conspiracy, but there was a transfer effect in critical thinking, which Gordon says is “a pretty unprecedented result in psychology.”

• Participants were more likely to try to convince other people out of conspiracy theories on social media after the chatbot conversation.

• Gordon’s team ran the experiments again and told participants that they were talking to a partisan bot — an AI model trained on Republican or Democratic data (even though it wasn’t). That had no impact. Being told the bot was biased did not weaken the effect of the facts and evidence it was providing at all.

• When researchers told participants that they were chatting with a live expert, and made it appear that way with a little spinny-wheel and all, that also had no effect. Whether they thought they were talking to an AI model or a human, the participants still changed their minds the same amount.

So what is going on here? Why does this work so effectively? Well, the main takeaway, Gordon says, is that facts matter.

I’m gonna say that again, for the cynics in the back: FACTS MATTER.

In later versions of the experiment, Gordon and his team controlled for different variables, and everything pointed to the notion that the thing that is changing people’s minds is the facts and evidence the AI is providing. And AI is uniquely suited to this task. It has the ability to quickly access arcane facts on a wide variety of topics. That would be hard for any single person to do, whether you’re a journalist, or someone sitting at a Thanksgiving table trying to debate your proverbial conspiracy-brained uncle.

Just think about 9/11. To comprehensively respond to someone who believes it was an inside job, you’d probably need a structural engineer to outline the causes of building collapse; a detonation expert to explain that explosives couldn’t have taken Building 7 down; a chemist who knows the properties of steel; an aviation expert; someone who knows the history of security measures in the U.S. after 9/11; and a flight-training expert. Conspiracy theories tend to sprawl in so many directions that almost no one has all the information at the ready to counter each point.

But being able to harness those facts and provide them to someone who believes 9/11 wasn’t a real outside attack can make a real difference.

“I think there’s this misconception about conspiracy theorists,” Gordon says. “They actually do care about the truth. If anything, they maybe care about it too much. That’s what led them down the rabbit hole. They’re trying to uncover some deep truth. And if you give them all this evidence, it’s gonna have some sort of impact.”

Now, the dangerous flip side of this is that information that appears to be factual can also be very persuasive in changing people’s minds. You or I may call this stuff lies, or propaganda, or disinformation. Gordon uses a term I’ve never heard before, but I get a kick out of: “ostensible facts.”

His research shows that AI bots using “ostensible facts” (disinformation) can also compellingly convince people of things. Which makes sense. And it’s a reminder that while there seems to be an opportunity to use this technology to create a more reliable, fact-driven internet, the opposite could also happen. AI can just as easily spread lies – perhaps in an even more insidious way than they’re circulated now.

But there is a silver lining: Gordon’s team ran another experiment in which they instructed the AI to convince participants to believe a conspiracy theory. But they told the model it had to use only accurate information. What they found: a significant amount of the time, the chatbot ended up defying its instructions. Once it realized there wasn’t accurate information out there to make the argument for the unfounded conspiracy theory, the bot took it upon itself to convince the participant of the truth.

Which, to me, just reiterates the lesson: facts do matter.

Facts matter. Facts matter. Facts matter. That’s what I take away from this fascinating research. Listen to this week’s episode where we dig into it more with Gordon Pennycook:

And try Debunkbot out for yourself, here, and read through the raw chats in Gordon’s experiment here.

Brian

Subscribe now

The episode is about how much false information AI chatbots are spewing to users, and how the most reliable chatbot, Claude, has started to get much less reliable in recent weeks — specifically, becoming more susceptible to Russian disinformation.

I interviewed an AI analyst from the fact-checking company NewsGuard, Isis Blachez, who walked me through one false claim: an allegation that Ukrainian border guards have been shooting hundreds of deserters each month in a river on the country’s border. The false claim appears to have traveled from a fake Human Rights Watch video, likely part of a Russian disinformation campaign, to a fake propaganda news site called Pravda, then into Claude’s model.

NewsGuard has been regularly testing the major chatbots for their trustworthiness, and Claude has consistently been the most reliable. According to their most recent audit in January, ChatGPT was spouting false claims about the news more than 30% of the time, and X’s Grok and Google’s Gemini did so about 40% of the time. Meanwhile, Claude didn’t repeat a single false claim in that audit. When asked about a piece of disinformation or a lie, it would correct the user, tell them it wasn’t true, and provide accurate information. Anthropic should be proud.

But in the last few weeks, something has changed. Users started complaining that Claude was getting less reliable. So NewsGuard ran a special test to see if it was true. And it is. Claude is suddenly putting out more bullshit — specifically, Russian disinformation. It turns out, as I discovered while reporting this week’s episode, that Kremlin-aligned propagandists have spent the last couple of years devising new forms of information warfare specifically aimed at manipulating chatbots into manipulating us. And while Claude was resistant to manipulation campaigns for a long time, it appears right now to be buckling. Anthropic told me it has policies and safeguards in place to stop Claude from being used to propagate misinformation, but they wouldn’t say whether they’re aware of the shrewd new tactics disinformers are developing to compromise chatbots like Claude, or if they’re updating Claude to resist them.

At the very least, I gave my friend Soph some ammunition for arguments with her stepdad! She forwarded him our episode with a warning: be careful what you read from AI.

Before I let you go: a request. It’s KCRW’s spring fundraiser. As many of you know, Congress defunded public media last year, which cost KCRW $1.3M in annual funding. That is essentially the cost of Question Everything’s budget, which goes entirely towards supporting independent reporting and journalism. In the last couple of months, the team at Question Everything and I have gotten at least a half-dozen tips about stories people want us to look into. One is about an artificial “news” site that appears to be fraudulently using the bylines of real reporters to peddle fake stories (and it might go far wider than just this one site); one tipster wants us to look into a very shady-seeming off-the-grid medical facility; another source has come to us with egregious allegations of corruption in local LA government related to last year’s fires.

Each of these stories is high stakes, in the public interest, and, importantly for us, illuminating about the state of the country and the ways people are trying to distort and bury the truth. But as with any tip, who knows if they hold up. Can we verify them? We need to pay reporters, fact-checkers, editors, and lawyers to figure it out. This is the work of fact-based, independent journalism, and it’s the work our industrious Question Everything team does every day.

People in power don’t want us to have the resources to do this work. The public media cuts were an attempt to diminish the ability of journalists to receive tips like the ones I mentioned, chase the truth, and deliver those stories reliably to you, the public. So we need your support. You can help us increase our capacity to do this work by becoming a monthly donor today, and when you do, your entire first year of giving will be DOUBLED, thanks to a generous $50,000 challenge grant from KCRW Board Member Michael Fleming and The David Bohnett Foundation.

Thank you,

Brian

P.S. Yes, as you may have noticed in the group chat screenshot, I do officiate weddings. If you’re getting hitched, give a shout.

Subscribe now

This is how a representative from Kalshi described prediction markets to my producer this week. I’ll be honest: I find the concept tantalizing. In the peak-capitalist inferno engulfing our country, where so much anti-truth activity is financially incentivized, the idea of fighting fire with fire is appealing.

Prediction markets, if you’re not familiar, are sites where you can put money on the outcome of an event. Will Idris Elba be the next James Bond? How long will the government shutdown last? What’s the temperature in Dallas going to be today? The companies that run them — the best known in the US right now is probably Kalshi — maintain that this is not gambling, and that they have tremendous value for the news and information ecosystem.

Believe it or not, there is some compelling evidence that supports this concept. For one, there was a study showing that people who bet on short-term indicators of climate change, such as sea level rise, came to believe more in the reality of climate change. The idea is that putting money behind real events motivates people to seek out reliable, factual information to base their decision on. Also, as one researcher theorized, what you make money on becomes part of your identity, so if you’re making money on evidence of climate change, you’re more likely to believe in it.

A lot of news orgs are now partnering with prediction markets, using their data the way they would polls, but also hawking the sites in the process. The AP, the Wall Street Journal, CNN, CNBC — even this platform, Substack — all have deals with prediction market companies now.

Despite the media industry drinking the Kalshi Kool-Aid , there are many reasons to worry about this expansion of prediction markets into news, and to expect that they might not be an overall positive for society. Not least of them: the impacts we’ve seen from the explosion in sports betting in the last eight years.

My team and I recently took over Bibber & Bell Wine and Spirits in New York after hours with three reporters who cover gambling and prediction markets to get the skinny on the rise of betting in sports, news, and American life. Kate Knibbs of WIRED, Danny Funt, who wrote the book Everybody Loses: The Tumultuous Rise of American Sports Gambling, and independent sports reporter Joon Lee shared their insights.

Here are just a few of the crazy things I learned from them:

➡️ The CFTC (Commodity Futures Trading Commission) is technically supposed to approve the bets on prediction markets. For example, when you’re placing a bet on Kalshi for which insulting nickname Donald Trump is going to say this month (will it be “Tampon Tim” or “Slopadopolous”?), apparently, a federal regulator approved that.

➡️ One thing that you won’t see trading on a prediction market: onion futures! Those are outlawed because of a scandal with a corrupt onion futures trader in the ‘50s (and a law called the Corrupt Onion Futures Act of 1958).

➡️ Kalshi offered a billion dollars to anyone who had a perfect March Madness bracket. The CEO, Tarek Mansour, encouraged people to enter the competition by saying it was an opportunity to build “generational wealth.... You owe it to your grandchildren.” The odds of getting every game right in a March Madness bracket? 1 in 9.2 QUINTILLION.

➡️ Betting companies are reaching out to strike deals with independent journalists. Two of the reporters who joined us recently got contacted by agencies trying to get them to refer customers. Here’s the email Joon got from a firm representing Kalshi very shortly after he criticized the company’s CEO on X:

➡️ I didn’t realize that the massive shift towards legalizing sports betting nationwide could be traced to a single op-ed by NBA Commissioner Adam Silver. Here’s Danny telling that story:

Listen to the whole fascinating conversation:

One last thing: I want to share the very thorough response we got from Kalshi media rep Jack Such when we reached out to the company with questions. I think it’s interesting to see the response in full because it gives some insight into how the people behind Kalshi see themselves, how they’re selling the company, and the points where they clearly feel on the defensive as people are expressing understandable worries about this new business that the CEO says is endeavoring to monetize “any difference in opinion.”

Oh and check out their response to our question about why they reached out to Joon with a partnership offer. Do I detect a bit of shade?

Brian

Subscribe now

The deadline is Thursday, April 16th. Vote here.

We’re nominated in the Podcast News & Politics category for the live episode we did last year at Tufts University, as ICE was detaining grad student Rümeysa Öztürk for an op-ed she co-authored in the student paper calling on the school to divest from Israeli-aligned investments. Core, constitutionally-protected speech that our government jailed her for. It’s definitely worth listening to, if you didn’t catch it at the time — we had legendary Washington Post editor Marty Baron join us, among other great guests.

And yeah, please vote for us if you’re so inclined.

Thanks!

Brian

Subscribe now

They had just attempted suicide. At 13.

At age 11, they signed up for Instagram using their brother’s birthday. Soon, Taylor says, Instagram sent them a recommendation to check out a stranger’s account. When they did, they found a diary of graphic self-harm photos. “It was gore,” Taylor told me.

Taylor remembers cutting themselves for the first time that night. Within days, they were drawn down a rabbit hole of self-harm and suicide content. They became an “everyday cutter.” By the middle of 7th grade, they were clinically depressed, suffering from an eating disorder, and suicidal. At 13, they tried to take their life. It wouldn’t be the last time.

Now, at 24, Taylor is doing better. No more suicidal ideation; no more eating disorder; and, crucially, no more social media. They quit. (These days, they use a bedazzled flip phone.)

Taylor is one of the thousands of young people now suing Meta and other social media companies, accusing them of stealing their childhoods. A couple of weeks ago, a jury in LA decided the first of these cases in favor of a 20-year-old named Kaley, finding that Meta and Google purposely designed Instagram and YouTube to be addictive to kids, knew the apps were very harmful, and pushed them on children anyway, without warning parents. Taylor could soon get their day in a federal court in Oakland.

Talking to Taylor was one of the most affecting interviews I’ve done in a long time. Taylor doesn’t solely blame social media for their struggles. They acknowledge other factors in their life that played into them. This is what Meta and other tech companies argue: that they shouldn’t be held responsible for the challenges of teenhood, and that social media addiction isn’t a real diagnosis. But Taylor described to me, very vulnerably, how their experience on social media — both with the content they were seeing, and the design of the app — exacerbated their IRL problems, and derailed them from getting the appropriate help. I also spoke to Taylor’s mom, Carissa, who reflected on her role and how innocuous Instagram seemed in 2013 before we knew how harmful it could be — brought to light largely from whistleblowers like Frances Haugen and these recent trials.

As someone who is Mark Zuckerberg’s age, my generation was the last to experience college before and after social media. In the past, I didn’t dismiss stories I’d heard about teens who said social media took over their childhoods and caused them to harm themselves. But I definitely didn’t understand them. I didn’t understand how a social media app could lead you to hurt yourself.

Until I talked to Taylor. Taylor helped me see the complex ways social media can embed itself into the lives, and day to day experiences, of teens.

I hope you check out Taylor’s story on “Question Everything”, and decide for yourself how much you think Instagram is to blame for what happened to Taylor.

And if you didn’t hear last week’s episode, it’s a good partner story to this one. I spoke to one of the lead attorneys who just succeeded in the case against Meta and Google in LA, Mariana McConnell, about the shrewd legal strategy they used to win.

Brian

Subscribe now

A 20-year-old in L.A. just sued two ginormous tech companies — Meta and Google — for intentionally designing social media apps that addicted her as a kid. And in a watershed verdict last week, she won.

One of the lawyers sent an email out to reporters right after the verdict was read aloud on Thursday. This was a civil case, so it was a list of questions the jurors had to answer for both Meta and YouTube.

It was a sweep — yesses all the way down, for both:

The jury decided the corporations have to pay the plaintiff, known in court by her first name, Kaley, or her initials, K.G.M., $6 million.

That’s probably lower than Meta and Google’s annual toilet paper budget. But when you think about the thousands of similar cases that have already been filed in court as part of a mass litigation against the companies, and the many more young people who could feel emboldened to sue now, if this bellwether case is a sign of how those could go, it definitely adds up.

“It felt great to get these answers,” co-lead attorney for Kaley, Mariana McConnell, told me shortly after the verdict. “ The jurors really understood that Meta and YouTube intentionally designed their platforms in a certain way to addict, and they did addict. And they addicted Kaley.”

As close followers of this newsletter and the Question Everything podcast will know, social media companies are almost always shielded from lawsuits by the law known as Section 230, which protects internet service providers from being sued for content other parties post (even if the providers moderate or curate it). So it’s consequential, well beyond Kaley’s case, that Mariana and her team were able to pull this off.

How did they do it? On our show this week, Mariana walks me through the shrewd legal argument she and her colleagues used to do an end run around Section 230.

The gist: they argued it as a product liability case. Not a free speech case. They said Instagram and YouTube’s design features were addictive and harmful to children. Not the content of the posts, but the architecture delivering them: blind notifications, infinite scroll, beauty filters, and of course, the algorithms driving the apps.

Some lawyers and writers argue that this is a distinction with no real difference. As Mike Masnick wrote in TechDirt:

This distinction — between “design” and “content” — sounds reasonable for about three seconds. Then you realize it falls apart completely.

Here’s a thought experiment: imagine Instagram, but every single post is a video of paint drying. Same infinite scroll. Same autoplay. Same algorithmic recommendations. Same notification systems. Is anyone addicted? Is anyone harmed? Is anyone suing?

Of course not. Because infinite scroll is not inherently harmful. Autoplay is not inherently harmful. Algorithmic recommendations are not inherently harmful. These features only matter because of the content they deliver. The “addictive design” does nothing without the underlying user-generated content that makes people want to keep scrolling.

But the judge in the case, Carolyn B. Kuhl, disagreed. She allowed Mariana and the legal team to pursue liability for design choices, as long as their case didn’t veer into third-party content.

 “The judge let us walk that path,” Mariana said, “without letting us, for example, put on the negative content that Kaylee saw on the platform. That wasn’t evidence in our case.”

 “Let’s take bullying. Our Case is not about bullies on the platform calling you names or harassing you. It’s not about adults and unwanted sexual contact by adults. It’s about the fact that even despite those things that might happen to you, despite the bullying that might happen to you on the platform, despite the fact that you could get creepy guys in your dms, you’re still coming back to the platform and you’re still using it — because you’re addicted to it.”

You can watch my full interview with Mariana here:

For their part, Meta and Google argued that social media addiction isn’t a recognized psychiatric diagnosis, such as alcoholism or gambling addiction, so it isn’t real.

But evidence Mariana’s team was able to extract from inside the companies certainly appears to undermine that claim. Case in point — this slide from an internal presentation from Meta:

Or this one:

Or a Meta doc that said of the lifetime value of users who start as children: “Young ones are the best ones.”

Meta’s chief legal officer, C.J. Mahoney, told Fox News: “We disagree with these verdicts, respectfully. We think they’re vulnerable on appeal, and we’re going to pursue those appeals aggressively.” Google also said it disagreed with the verdict and would appeal.

It’s likely they’ll appeal on Section 230 grounds, arguing that the judge was mistaken in letting the case move forward in the first place. That it never should’ve even been heard in front of a jury.

And it’s not impossible that argument could win. Courts have traditionally read Section 230 — and the First Amendment — to give very broad protections to internet service providers.

But for now, this verdict stands. A jury reviewed the evidence and testimony for months, deliberated for eight days, and decided that two of the biggest companies in the world, which for years have hidden behind a free speech shield, were hurting kids for profit. And it’ll be hard to put that genie back in the bottle.

Brian

P.S. Stay tuned for next week, when I share one of the most affecting interviews I’ve done in a long time, with one of the young people who is in line after Kaley to sue the social media companies.

Subscribe now

Evidence from the trial revealed that Meta knew Instagram’s algorithm was recommending nearly four times as many minors to accounts flagged for “groomer-esque” behaviors (adults following teen hashtags, or so-called “sexy teen accounts”) than it did to ordinary adult users. According to internal documents, 27 percent of the accounts recommended to these likely groomers were minors, as opposed to 7 percent for regular adults.

I found this enraging for a bunch of reasons. First, according to The Atlantic reporters, Instagram didn’t lock down accounts for its youngest users until years after learning these statistics. But also: there’s an argument that algorithms are a form of protected speech, and that regulating them, or carving them out from Section 230’s immunity, would violate the First Amendment. An algorithmic recommendation, the argument goes, is just a machine’s – or a company’s – opinion about what a user might like to see. Under that logic, the government can’t squash it.

Really? Is a machine-learning algorithm that connects children to potential sex abusers First Amendment-protected speech? I love the First Amendment. But I don’t think so.

This is the thing I tried to hammer home in a debate I was part of yesterday on the YouTube/radio show, The Majority Report (you can watch it below). Hosts Sam Seder and Emma Vigeland had me on with a critic of mine: longtime tech writer Mike Masnick. He founded the blog, TechDirt, created his own podcast about Section 230, and is a big defender of the law.

When we first set our sights on Section 230 here at Question Everything, Mike came out with a pretty tough rebuttal of our first episode about it. He basically claimed I didn’t understand the law. I find this to be a somewhat common refrain among Section 230 fans – they tend to respond to criticism of the statute by saying the person criticizing it just doesn’t understand it. Tech reporter Taylor Lorenz is another Section 230 defender who did this to me yesterday:

I’m not sure what the whole “This Brian Reed guy” thing is about: Taylor knows me, I’ve interviewed her for another story (it hasn’t run yet), and we’ve emailed about Section 230. But anyway… As I responded to Taylor, it is possible to understand Section 230, care a ton about the First Amendment (as I do), and still believe Section 230 should be revisited.

It’s not as if once we all reach some immaculate understanding of Section 230, we will see that it can never be criticized or re-envisioned. Even the most OG Section 230 expert and booster, Eric Goldman of the High Tech Law Institute, calls Section 230 the “least worst” option. There are tons of downsides to Section 230, and it seems crazy not to dig into them and ask whether there might be a better framework for the modern internet.

So I’m glad that is what we did, pretty substantively, on The Majority Report.

You should watch if you’re interested. The big point I was trying to make: Section 230 is supposed to be a “Good Samaritan” statute. That’s actually in the text of Section 230. Internet service providers – apps, platforms, blogs, etc – are shielded from lawsuits over stuff users post on their site, even if those platforms moderate, block, and or amplify that content. The law is supposed to encourage good behavior from companies, creating a positive experience for users.

But it doesn’t require it. Companies get the immunity whether they’re good Samaritans or not. And I believe the evidence shows that the most powerful companies in tech have not lived up to their end of the bargain. They have not proven to be good Samaritans. The revelations about Instagram feeding kids’ accounts to potential groomers are just the latest awful example. Yet Meta and other tech companies have profited mightily from the liability shield they get. That’s why the law warrants another look.

For Mike’s part, I thought the most compelling points he made had to do with the Computer Fraud and Abuse Act, which he believes is the main culprit for why the Internet sucks today (and not Section 230). And also about the effect copyright takedown requests have had on competition in the video streaming world – why, really, we only have YouTube.

As always, let me know what you think, what questions you have, what points you think I’m missing. And if you have tips – write me here, or on Signal at brihreed.45

Brian

Subscribe now

What a week.

Question Everything won an Ambie Award from The Podcast Academy on Tuesday — this is like the Oscars of podcasting — for Best Reporting. The winning story was executive producer Robyn Semien’s two-part series about Jeremy Loffredo, who at the end of 2024 became the first American journalist ever arrested by Israel. It was produced by Zach St. Louis, and edited by contributing editor Joel Lovell, among others on our team.

If you didn’t catch these episodes when they came out last January, go check them out now.

Part One:

Part Two:

What's so remarkable about these episodes is that Robyn started telling Jeremy's story because what Israel did to him was so messed up. But once she started investigating, she found that Jeremy, and the outlet he worked for, The Grayzone, also did some messed up stuff. It didn't justify what Israel did to him at all. But it was important to shine a light on, even if Jeremy and The Grayzone didn’t like us doing that.

It's a story that challenges easy lines of thinking, which is what we're always striving for with our show. It’s awesome that a group of our peers agreed when they voted.

In addition to Robyn and Zach, I'm so proud of our whole team at Placement Theory, and our partners at KCRW (including Arnie Seipel, Jennfer Ferro, Tejal Ajmera, and Nathalie Hill) for supporting this kind of ambitious, and award-winning (!) reporting.

Robyn and Zach couldn’t make the awards in Brooklyn, sadly, due to the blizzard, so I did my best to represent with our contributing editor Neil Drumming.

We have another big two-part investigation that we just dropped: the story of Mario Guevara, a journalist arrested last year…in America. Jailed for his reporting.

You can hear those episodes, reported by Sophie Kazis, here - Part One, Part Two - they’re excellent.

I promise not all of our stories will be about journalists getting arrested but unfortunately this is a growing activity of oppressive governments right now, and you know what they say, about first coming for the journalists…and then…

Brian

P.S. If you haven’t already, please take 5 minutes to fill out this survey about Question Everything and let us know what you do or don’t like about the show. It helps us help you. Thanks.

Subscribe now

Mark Zuckerberg didn’t want to be there. And it showed.

“ I was frankly a little surprised with the Mark that we got,” said Sonja Sharp, legal affairs correspondent for the L.A. Times, who was sitting in the courtroom during Zuckerberg’s testimony this week. I talked to her the morning after Zuckerberg finished testifying to get the readout.

“He was much more combative than I had expected him to be.”

A judge ordered the Meta CEO to testify in what’s shaping up to be a landmark “social media addiction” trial in Los Angeles. A 20-year-old woman, going by her first name, Kaley, is suing Meta and Google (after settling with Snap and TikTok), accusing the companies of creating a social media addiction that led to depression, anxiety, and suicidal thoughts. Hers is the first of roughly 3,000 scheduled to be argued in Los Angeles County Superior Court.

For years, Meta has been able to avoid these kinds of lawsuits thanks to the protections of Section 230 — the law that shields internet companies like his from liability for content posted on their site.

But in response to a novel legal strategy by Kaley’s attorneys, this judge carved out an opening: the companies can’t be sued for what users post, but they can be sued if they designed their products to addict minors. If it’s a product liability case, and not a speech case, Section 230 can be circumvented.

This is the first time Zuckerberg has testified in one of these cases — he tried to get out of it, but the judge forced him to appear — and he acted very Zuckerberg-y.

“I think I’m sort of well-known to be pretty bad at this,” Zuckerberg admitted on the stand.

“[Zuckerberg] was, I would say, fairly pedantic at times, which didn’t play well with the gallery,” Sharp said.

At one point, Zuckerberg even refused to agree that the date he had testified before Congress was January 31, 2024. At another turn, he hesitated to concede that if something is addictive, people may “do it more.”

It was a contrast to the more stilted, apologetic version of Zuckerberg many remember from that Congressional testimony, when Republican Senator Josh Hawley of Missouri forced him to turn around and apologize to families behind him for the torment his company’s platforms had caused their children. This time, his tone sounded more contentious.

The plaintiff’s attorney, Matt Lanier, brought up that past sworn testimony and Meta’s internal documents. While Zuckerberg had claimed that children under 13 are not allowed on Instagram, Lanier showed internal 2015 documents estimating that there were then four million Instagram users under 13.

Lanier also presented internal goal-setting emails from 2015. In one, increasing time spent on the platform was listed as the first item.

Zuckerberg pushed back, arguing that the metric wasn’t a “goal,” but a “milestone.”

Pedantry.

“I think that he may have lost some folks in being really emphatic that these incredibly fine-grained distinctions in business terms were meaningful,” Sharp said.

Lanier presented emails suggesting that Zuckerberg had some familiarity with his platform’s attempts to keep children engaged.

“The Meta employees in these documents are saying, ‘This is coming from Mark. Mark wants us to prioritize teens. Mark wants more time spent on the sites.’” said Miles Klee, senior culture writer at WIRED, who was present during the testimony. “ You see all these internal communications from people who work there, including people who are kind of disgusted by the idea that he wants to get under-13-year-olds addicted to this stuff.”

Zuckerberg attempted to refute many of the emails presented as evidence based on how old the email was, or he claimed that they involved Meta employees that he was not familiar with.

For instance, when asked if he knew Karina Newton, Instagram’s head of policy in 2021, he replied, “I don’t think so, no.”

Even as the questioning became increasingly heated, the mood in the court was far more reserved than during Instagram head Adam Mosseri’s testimony last week, Sharp told me.

She said parents who blame social media for teen suicides started sobbing when Mosseri alleged that social media could not be “clinically addictive” and that 16 hours of daily social media usage is not “addiction.”

“Families were breaking down in the court to the point that the judge admonished them,” Sharp said.

Meta’s defense has focused heavily on Kaley’s medical records and personal history, arguing that her struggles predated and cannot be pinned on product design. Sharp expects them to continue this argument throughout the trial.

“They really want to hammer home both these almost contradictory ideas, first of all, that social media addiction is not a real thing. And then second of all, she doesn’t have it,” Sharp said.

Another interesting thing Sharp mentioned: changes to the jury may pose a significant problem for the defense, as one juror, an older man, was replaced with a middle-aged woman. Young and middle-aged women are a demographic seen to be more favorable to the plaintiff, Sharp said.

Also this week, former Meta executive Brian Boland testified. According to The Verge, he said that over his 11 years at Meta, he went from having a blind faith in Facebook to feeling that Zuckerberg cared most about “competition and power and growth.”

Subscribe now

Something unusual is happening in a Los Angeles courtroom that could change the way social media companies do business.

For decades, Section 230 — the 1996 law that gives immunity to internet service providers for what their users post — has acted like a legal force field around internet platforms. Thousands of lawsuits have been dismissed before they reached trial. Cases were even thrown out where families argued that social media platforms’ recommended content were to blame for overdose deaths and terrorist attacks against their loved ones. Courts ruled that Section 230 still protected the platforms.

But now — perhaps fittingly, on the 30th anniversary of Section 230’s enactment — a Los Angeles judge has allowed a case against social media companies to go forward, despite the law’s protections. Opening arguments started on Monday.

A 20-year-old woman identified as K.G.M., or Kaley, is suing Meta and Google. She’s the first plaintiff in a line of roughly 3,000 others who are trying to hold the tech companies accountable in Los Angeles Superior Court. Her case is considered a bellwether trial, which will test how juries are likely to rule on the thousands of other claims, and could spur settlements if the plaintiff wins.

Kaley says she began using YouTube at six and Instagram at nine, and claims the platforms hooked her into a social media addiction that led to depression, anxiety, body image obsession, and suicidal thoughts. Her lawyers are arguing that the social media products are designed, similar to slot machines, to be addictive, and that the companies know this.

Snapchat and TikTok have already settled with Kaley. The terms haven’t been disclosed, but the plaintiff’s legal team told us the agreements were financial, and the companies did not agree to make any changes to the way they operate.

Meta and Google might not get off so easily.

Subscribe now

Puncturing Section 230

Leading up to the trial, Google (which owns YouTube) argued that Kaley’s legal team should be barred from introducing evidence about how YouTube is designed, including how the platform collects behavioral data and recommends content. They said Section 230 should protect them from having this evidence used in court.

But in a big ruling last weekend, Judge Carolyn Kuhl said no.

She ruled against Google, saying that Section 230 “does not prevent Plaintiffs from holding Defendants liable for using data collection to make their platforms addictive to minor users regardless of the specific content that is shown.”

In other words, you still can’t sue a platform for something a user posted. But, in this judge’s view, Section 230 does not automatically shield companies from liability when it comes to how their systems are set up. So, if the companies design their product in a way that feeds a harmful addiction — with tools like autoplay, likes, algorithmic recommendations — you can try to sue them for that.

It might seem like a subtle difference. But it’s a crucial one that is allowing Kaley to actually have her day in court. A rarity in the U.S., because of Section 230.

Traps Not Apps

Our managing editor, Kevin Sullivan, spoke this week with Mariana McConnell, co-lead counsel for the plaintiff, and she framed their strategy this way: “Traps, not apps.”

She said their case will rely heavily on internal documents, including employee chatter in which one Meta staffer wrote, “Oh my gosh yall [Instagram] is a drug,” and another described employees as “basically pushers.”

Meta’s internal “Project Myst” study, she said, found that teens who had experienced trauma were especially vulnerable to compulsive use, and that parental controls “do nothing.”

This was discussed in the courtroom this week, and I found it pretty remarkable: there’s a line of argument that it’s parents’ responsibility to monitor their kids’ social media use. Mariana told Kevin that the defense will try to blame parents. But Meta’s own research shows that parents face extensive barriers in trying to curb kids’ social media use.

“[Kaley] really stood no chance against these big tech companies,” Mariana said.

In opening arguments on Monday, lead trial lawyer Mark Lanier told jurors the case was “as easy as ABC” — short for “Addicting the Brains of Children.” He compared YouTube to a casino that hooks gamblers with slot machines. And he drew parallels to Big Tobacco, arguing that internal research shows that these companies knew they were harming children, but they did it anyway.

We reached out to the legal teams for Meta and Google for comment. Meta pointed us to a website they’ve designed that says the company has been taking steps to protect minors, with things like the introduction of “Teen Accounts,” which they say “have built-in protections that automatically limit who can contact them and the content they see.”

In court on Wednesday, Adam Mosseri, the head of Instagram (owned by Meta), testified that the company has invested heavily in youth protections and is trying to balance safety with free expression, saying they aim to be “as safe as possible but also censor as little as possible.”

Share

Emotional Courtroom Scenes

The fact that a top tech exec like Mosseri even had to testify in court is a big deal. And his testimony prompted emotional reactions in the courtroom.

I spoke to Reuters technology policy reporter Jody Godoy, who was in court watching Mosseri. The courtroom was packed, she said, with many parents who say social media contributed to their children’s suicides and self-harm. At one point, Mosseri was testifying about so-called “beauty filters” — a popular feature that lets people change their appearance to plump their lips, slim their nose, or change the shape of their eyes. They have been blamed for causing body dysmorphia and other emotional turmoil in teens. Jody noticed how one mom responded to his testimony.

“She started to visibly cry and have to wipe her eyes because her daughter… died when she was 16. She saw the effect that those types of filters and that type of content had on her daughter,” Jody told me.

“[The mom] thought that the way that [the filters] were being discussed in court was cavalier, and she said outside court: ‘Our kids were guinea pigs in this move fast and break things culture.’”

“Move fast and break things” was an early mantra of Facebook founder Mark Zuckerberg: an edgy slogan when you’re a startup full of dudes in your twenties. Not so edgy when you control some of the most powerful communications platforms in the world that are used by millions of children.

A Wave of Cases

The companies argue that they cannot be held liable for speech created by billions of users, and making them liable for the way they recommend content would fundamentally alter how the modern internet functions.

Now a jury is being asked to decide something different. Whether these companies created an environment designed to keep children hooked.

This case is the first in a tidal wave of similar cases that could bring social media giants to their knees.

Three thousand lawsuits are lined up in California Superior Court behind Kaley’s.

Also this week, opening statements began in New Mexico, where the state attorney general alleges Meta failed to protect minors from sexual exploitation.

The judge in that case dismissed Meta’s attempts to invoke Section 230 and the First Amendment, ruling in June 2024 that the case focuses on platform product design and other issues not related to free speech.

During opening arguments this week, Meta’s lawyer said the company is working to protect everyone, including children.

“Meta works incredibly hard to protect its users. Meta has built amazing automated tools to protect people. Meta has 40,000 people working to make Facebook and Instagram as safe as possible,” he said.

The New Mexico case is one of more than 40 lawsuits from attorneys general across the country.

And in June, the first of many federal cases on behalf of school districts will begin in Oakland, California. The schools claim social media has harmed children’s mental health and forced them to invest in more resources to deal with the crisis. The tech companies tried to stop the case by invoking Section 230, but the judge didn’t go for it. But she did say there would be lengthy debate about what could be shown to juries during the upcoming trials.

Mariana told Kevin that she and many of the lawyers on Kaley’s case came of age alongside social media.

“We were probably the last generation to kind of get out alive,” she said.

She also said she expects the tech companies to keep trying to use Section 230 throughout the trial to make it harder for her client to win. But the fact that they’re even having a trial — where next week, Mark Zuckerberg himself is expected to take the stand — is a huge turning point.

Subscribe now