But that isn’t what Kennedy believes. It’s not what he demonstrates through actions, either. And if you want proof of that, you need only read about how Kennedy has unilaterally adjusted rules at HHS to allow pharmacies to sell you peptides over the objections of the FDA.

Anti-vaccine Health Secretary Robert F. Kennedy has unilaterally passed a rule that will allow specialty pharmacies to sell six specific peptide drugs, overruling scientists at the Food and Drug Administration.

Kennedy’s new rule sidesteps the need to formally undo that FDA action. The rule is titled “Enforcement Discretion Policy Regarding 503A Compounding Using Certain Peptide Bulk Drug Substances” and was made available Wednesday on the Federal Register, with the final publication date set for October 30. The rule does not change the regulatory status of the peptide; instead, it states the Department of Health and Human Services (HHS) will no longer take enforcement action against compounding pharmacies for making the peptides, essentially giving them a green light to freely produce and sell them.

So what qualifications does Kennedy possess that is informing his decision to overrule actual FDA scientists on the topic of peptides? None, of course. Kennedy is not a scientist. He’s not a healthcare professional. What he is, instead, is both a user of peptides personally and someone who has many friends in the peptide business. It’s just another grift, though it is being wrapped in pretend backing from the panels that Kennedy personally stocked with sycophants.

One reason cited for all the peptides was that an FDA advisory panel voted in July to add these six peptides back to the list of drugs compound pharmacies can make. However, prior to the vote, Kennedy stacked that 12-member committee with eight new members. At least six of Kennedy’s additions have ties to peptide-related businesses and clinics, raising serious conflict-of-interest concerns.

Another explanation for allowing the peptides is that the safety concerns are mainly hypothetical or “inferential” and not based on observed harm in humans. But this is mainly because there are few to no studies involving humans that show either benefits or harms from the peptides.

And here we put the lie to Kennedy’s concerns about vaccines. If he wants to claim that vaccines haven’t been tested enough to be in wide public use, he cannot create a non-enforcement policy over peptides that have had far less, if any, such testing. We are living under health policies that are purely Kennedy’s personal whims. That’s nothing like the gold standard science that he has been promising to follow throughout his tenure at HHS.

The peptide pushers are celebrating and it’s not hard to understand why. Many Kennedy allies are among them. He gave them a gift, almost perfectly circumventing his previous fears about testing to do so.

One overarching point is that what the plaintiffs are trying to do is not what copyright law is for. Copyright gives the holder the ability to control how the embodiment of the work is used (ex: you can’t copy a book without permission, or, as the example we used in the brief, perform a play) but not the ability to control the public from benefiting from its contents. Copyright can control such activities as copying, distributing, or performing a work but NOT reading, watching, listening to, or learning from a work. What OpenAI does with its training is the latter set of activities, not the former. But if OpenAI is somehow not able to engage in these activities because of copyright, then no one can. Which would lead to a direct harm to the public, a loss of an important benefit copyright law is actually supposed to give it.

In fact, it is a benefit copyright law must provide the public. The whole point of copyright law is to promote “the progress of sciences and the useful arts,” or, in other words, help ensure we get more innovation and expression. But the plaintiffs in these cases want the opposite: they want copyright to stop these things. They want copyright law to prevent AI from being further developed (indeed, in its complaint the New York Times Company even asked for OpenAI’s models to be destroyed!) by denying it the ability to do the learning that is needed for it to be useful. And they want to prevent any of the further expression that AI helps facilitate, even though fostering new expression is exactly what copyright law is supposed to do.

In fact, it is what copyright law is constitutionally required to do. Moreover, if, rather than fostering innovation and expression, copyright law instead obstructs it, then copyright law is not constitutional. Congress only gets its authority to make copyright law through the Progress Clause (“to promote the progress of science and the useful arts…”) so if it turns out that the law it has passed instead does the opposite of promoting progress, then that law is unconstitutional, as being beyond its authority.

The First Amendment also limits Congress’s authority to pass laws that inhibit free expression. So Congress also would not have had the authority to pass a law that let copyright holders of existing works prevent either the development of AI or any of the downstream, lawful expression AI helps facilitate. If the copyright statute Congress passed were to have that effect, then it would be unconstitutional.

It also would violate the Constitution in another way we’ve discussed before: by violating the First Amendment-protected “right to read,” which is sometimes also referred to as the right to receive information. All those learning activities (reading, watching, listening to, etc.), which the current statute does not actually restrict, are activities that the First Amendment entitles the public to do, and copyright law not only does not prevent, but constitutionally could not prevent.

Meanwhile, we also made an overarching point: even to the extent that AI as we know it may be terrible, and we are looking for a law that can stop it, copyright law is still not the right law to do it. First of all, it’s not tailored to the actual problems we’d want to address (for instance, if using AI in hiring decisions produces discriminatory results because of bias incorporated into the AI model, then the solution to that problem is making sure that AI can train on even more material, not less, because it is the training limits that tend to create that bias). Secondly, the worse we think today’s AI is, the more important it is that something better is developed, but if law can stand in the way of further development, then none will be. None can be, and we’ll be stuck with the technology we have, if not an even worse version of it if we have to degrade the models in response to copyright claims. As well as stuck with an imbalanced, super-powered copyright law that could stand in the way of lots of other non-AI innovation and expression.

And, again, if we have to freeze the technology because of copyright law, then copyright law is definitely running afoul of the Constitution, which allows Congress to pass a copyright law that only does the opposite.

In sum, the worse you think AI is—and we make no quibble with that assessment, although we did point out that it also has some distinct upsides, particularly when it comes to enabling more expression—the more important it is that these copyright claims be dismissed. Because however bad the state of play may be today, a copyright law that would let these claims prevail will only make everything worse.

Subscribe now on Apple Podcasts, Overcast, Spotify, Pocket Casts, YouTube, or your podcast app of choice — or go straight to the RSS feed. To get extended episodes with additional coverage, support us on Patreon.

In this week’s episode, Mike and Ben cover:

• Lula’s Win Signals Fight to Regulate Social Media Will Escalate in Brazil (TechPolicy.Press)
• Meta Employees Denounce Interference in Brazilian Elections (Panamerican Dispatch)
• Google’s AI Chatbot Refuses to Answer Question of Who Won 2020 Presidential Election (Mediaite)
• AI-driven disinformation running rampant on social media before US midterms, experts warn (The Guardian)
• Never Mind The Deepfakes, Chatbots Have To Deal With Election Lies Being Free, While Truth Is Paywalled (Techdirt)
• French Government Calls Emergency Meetings With Platforms Over Protests (Wired)
• Farewell, France (Quad9)

And in the extended episode for Patreon supporters, they cover:

• Ofcom’s Fix For An Online Safety Act That Isn’t Working: More Online Safety Act (Techdirt)
• How age checks at age 16 can work to underpin UK’s social media ban (Ofcom)

Our fun links this week include Krillion (much harder than it looks) and Dial Flipper for discovering new podcasts.

Follow us on Instagram, YouTube, and Bluesky for video clips from the episode!

If you’re already a Patreon supporter, you can get the extended episode on Patreon.

This year alone, he’s issued Wyden Siren letters a few different times. And now he’s done so again, once again about Section 702 of the FISA Act, which has been a Wyden target for years. Wired has a summary:

Two Democratic senators who have long overseen US intelligence say the Trump administration is illegally withholding a secret court’s findings from Congress until lawmakers renew a program that lets federal agents read Americans’ communications without a warrant.

Seven months ago, the Foreign Intelligence Surveillance Court objected to search filters that let intelligence analysts pull Americans’ messages from intercepted data without the approvals, written justifications, or records the law requires. The court approved another year of surveillance anyway. The ruling has never been made public, and Senators Ron Wyden of Oregon and Adam Schiff of California say the administration’s own deadline for releasing it passed in September.

The program, Section 702 of the Foreign Intelligence Surveillance Act (FISA), lapsed in June after Congress failed to renew it, though collection continues under the court’s March approval. The administration’s own public documentation says rulings like this one are reviewed for public release “within 180 days of issuance.” That window closed September 13, even as intelligence officials press lawmakers to reauthorize the program by the end of the year.

Wyden sounding the alarm is expected (and welcome). Schiff co-signing it is, however, a surprise. For many years when he was a Representative (prior to recently moving over to the Senate), Schiff was a reliable voice and vote for greater surveillance powers. He would frequently roll out his “as a former prosecutor” line to defend giving more and more power to the government.

Indeed, back in 2018, we called out Schiff (who was then the House ranking member on the Intel Committee) for giving Donald Trump greater surveillance powers in renewing Section 702. It was bizarre: Schiff was one of the loudest voices calling out Trump’s abuses of power during his first term, and yet he led the charge among Democrats to hand Trump even more surveillance power. He even undercut Wyden’s own attempts to rein in those powers, as Trevor Timm noted at the time in an op-ed blasting Democrats for handing Trump more surveillance power:

Schiff was a leading driver in the House to extend the NSA’s surveillance powers, and has been undercutting the more robust reforms proposed by other Democrats, like longtime Senate Intelligence Committee member Sen. Ron Wyden, for months.

Given that history, seeing Schiff co-sign a letter accusing the administration of illegally hiding a ruling about 702 abuses is quite the turnaround.

Indeed, Schiff even blasted out a press release talking about “serious abuses of American’s rights.” That’s quite a statement given how strongly Schiff fought to give those powers to the government.

The specifics here involve a secretive FISC ruling regarding Section 702. Technically 702 expired back in June, after Trump tried to tie its renewal to his ridiculous voting bill that no one is willing to pass. But back in March there was a secret ruling by the FISA Court which allows the government to keep using this surveillance authority at least until March of next year. That same ruling also objected to the search filters analysts use to pull Americans’ messages without the approvals and records the law requires — and then approved another year of surveillance anyway. Wyden and Schiff want it public, and the law is on their side: the statute requires either a declassified version within 180 days or if the DNI waives declassification by yelling “national security” then it requires an unclassified statement from the Attorney General summarizing the legal interpretation. And Attorney General Todd Blanche unequivocally promised the Senate Judiciary Committee that he would deliver the unclassified opinion by the deadline.

That deadline came and went on September 13. The ruling remains classified. From Wyden and Schiff’s letter:

We and others have repeatedly called for the administration to release the March FISC opinion. In May, the Chair and Vice Chair of the Senate Intelligence Committee sent a letter to the administration stating their expectation that the opinion would be released by May 15 “in order to inform the Senate debate on section 702.” The statute itself requires the administration to release a declassified version of the opinion within 180 days. In July, Attorney General Blanche told the Senate Judiciary Committee he expected that the public release would happen “in advance of the statutory deadline.” Attorney General Blanche unequivocally committed to “declassifying the opinion by the statutory deadline of 180 days and complying with the law.” Nevertheless, it is now a month past the statutory deadline and the opinion has still not been released.

Simply put, this is not acceptable, especially as this administration restarts its attempts to lobby Congress to renew Section 702 without reforms. The administration has now had over seven months to declassify and release this opinion. It is time to stop violating the law and finally release this opinion. And if this administration won’t follow this simple statutory directive, it is unclear why Congress or the American people should trust this administration to follow any of the other legal limits to Section 702.

We demand that you release this opinion as soon as possible. We also request that you respond in writing by November 6, 2026, providing an explanation as to why the opinion was not released by the statutory deadline and identifying the officials that were responsible for withholding its timely release.

As Wired notes, Senators can read the ruling, but only in a secure room, and those who have done so are dropping heavy hints about what it reveals:

Lawmakers who have read the ruling hint that it says more. Senator Dick Durbin, the top Democrat on the Judiciary Committee, urged colleagues on the Senate floor in June to read it in a secure room. “I have seen it,” he said. Once they did, he told them, “your opinion of the role of FISA and the rights of Americans will be impacted for sure.”

Wyden has also warned about a classified interpretation of Section 702 that he says multiple administrations have relied on and refused to declassify. “When it is eventually declassified,” he said in March, “the American people are going to be stunned that it took so long and that Congress had been debating this authority with insufficient information.”

Of course, the Trump administration seems to believe that laws are meaningless, especially if they require transparency or limit the government’s ability to abuse the rights of anyone in America. So don’t hold your breath waiting for Director of National Intelligence Jay Clayton to cough up this ruling. But as Wyden and Schiff point out, if the administration can’t be bothered to follow a simple disclosure deadline, why should anyone believe it’s respecting any of 702’s other limits while it begs Congress for a clean renewal?

Note: The Techdirt Deals Store is powered and curated by StackSocial. A portion of all sales from Techdirt Deals helps support Techdirt. The products featured do not reflect endorsements by our editorial team.

First, there’s the regime change. Trump has been stupid as fuck all the time since day one but nothing was much stupider than issuing an executive order renaming the Gulf of Mexico as the “Gulf of America.” Oh wait. There was more stupidity. Trump — again saying anything that momentarily glanced off the shiny, wrinkless lobes of his forebrain — demanded Lake Ontario be renamed “Lake America.”

If that were the end of the story — a wet-brained spray-tan-maxxing fuckmook participating in some short attention span jabbering between hours of Truth Social doomscrolling — we wouldn’t be here today discussing this latest turn of events.

But Trump said something and companies started doing something. Some even started doing something before the spittle dried on Trump’s lips, as Karl Bode discussed earlier:

Google was the first to quickly make the change to appease Trump; in fact they acted so quickly on this the change to Google Maps happened before the government had even finished implementing it. Apple was very quick to follow suit, despite the fact that nobody at the company actually supports the ridiculous and pointless change.

That’s how tech billionaires greeted this momentary outburst from a guy who’s never bothered to think twice. One company altered its map before it was even made official. The other did it even though no one in the company actually wanted to have it done. These are the actions of two of the richest and most powerful companies in the world — companies whose fortunes would have dipped to the equivalent of a rounding error if they’d chosen to reject this overture.

Meanwhile, one of the first names to the (map) game refused to play along. MapQuest doesn’t have billions of revenue to spare, much less the ubiquity of companies that offer phones as well as pre-installed map apps. MapQuest once owned the digital mapping world, but it has been sidelined by vertical integration.

Maybe it’s just because MapQuest (relatively) has nothing to lose. Or maybe it’s because MapQuest actually seems to care about its users. Either way, MapQuest refused to alter its map to reflect Trump’s digital takeover of Lake Ontario.

And now MapQuest is going after America’s surveillance state — one made most infamous by another company that has far more liquidity than MapQuest: Flock Safety. While it’s certainly not the first company to offer automatic license plate readers, it’s the most famous. Its network is huge, capable of capturing 20 billion plate reads a month. And it’s the current favorite of law enforcement agencies all over the US.

But Flock is working against its own ubiquity at the moment. It’s more famous now for allowing cops to stalk people, especially ex-wives and girlfriends, as well as any people their exes might now be seeing. It’s also a handy tool for cops who just want to track people, including anti-police violence protesters or anyone else who’s annoyed them.

People want to know where these cameras are. Flock’s tools have generated a ton of public records that have ultimately given activists the ability to locate and map camera locations. Other companies working with law enforcement have generated some data as well, albeit not as much as the current market leader.

MapQuest now offers what other map apps don’t: a handy map of ALPRs.

MapQuest users in Canada and the U.S. can now see where some surveillance cameras have been set up in their communities.

The company introduced a “surveillance layer” earlier this week that uses animated eyes to mark where automated licence plate readers, or ALPRs, have been set up by police or others.

“We knew that there were a ton of cameras in use and documented in the U.S. and Canada,” MapQuest general manager Doug Berger told CBC News. “You would never know unless they were pointed out to you.”

This is very cool. For multiple reasons. A company deciding to highlight law enforcement surveillance on its official app is completely unexpected. That it did this because some of its users asked it to is even better, which means MapQuest cares more about its (relatively) small number of users than Google and Apple care about their millions of users who default to their apps simply because its part of their smartphone’s pre-installed ecosystem.

And that means MapQuest will continue to grow, three decades after its debut. GM Berger noted that the company gained three million new users after refusing to rename Lake Ontario. Adding a “surveillance layer” just means more deliberate interaction from users. And people who choose apps for a reason, rather than just using whatever first presents itself to them, are far more loyal.

This is a very smart move by MapQuest. The decisions it has made in the last couple of months shows it’s far more respectful of its current and future users than the massive tech companies that assume (often correctly) that consumers will eat whatever shit is shoveled at them by their service providers. MapQuest is looking for people who want to use MapQuest because they like what the company is doing. It’s an incredibly smart move that builds the sort of trust and loyalty larger companies can only wish they had.

So instead we get corruption-fueled protectionism run by a bunch of racist, and incompetent clowns who gutted most of our federal regulators, then saddled them with an elaborate new policy ask that’s impossible to implement, even if the admin were competent.

Which it isn’t.

For example, I recommend this recent Verge piece about the Trump effort to ban Chinese drones, which notes how it has been trivial for Chinese companies to trick the administration and bypass the ban simply by renting a strip mall office, tweaking a few names, and pretending their equipment is produced domestically. Curiously, officials who don’t believe in government can’t govern. Fancy that.

Another report this week by Ars Technica indicates that U.S. chip giant Intel has also been so disoriented and confused they felt it necessary to get a waiver from the Trump administration’s “foreign robot ban” because executives didn’t really understand what is or isn’t covered:

“Intel’s waiver request stated that it was “unclear” whether its robotic vehicles even qualify as robots covered by the FCC ban but said it sought relief just in case. The tech company also received approval to “make certain hardware, software, and firmware changes” that would help ensure the continued operation and maintenance of the robotic vehicles.”

So it’s a foreign robot ban, unless we need foreign robots. In which case big U.S. companies have to spend a bunch of time and money kissing the Trump administration’s ass to ensure they have what they need to function. It’s far from the “free market” dogma the administration rose to power on. It’s a weird and racist autocratic nanny state operated by the dimmest people imaginable.

A recent study by the International Data Corporation predicts that the foreign robot ban could reduce US humanoid robot revenues by 58 percent over expected baseline. China is still expected to dominate the market with most of its cheaper, better gear avoiding the U.S. for the EU. And much like the FCC “drone ban,” a big impact will be less tech product availability and higher prices overall for U.S. consumers.

Not content with screwing up the drone, robot, vacuum, and lawnmower sectors, the Trump FCC last week announced it would also soon implement a ban on “foreign surveillance chips.” You’re to ignore that the FCC can’t really define what this means. And doesn’t really want to answer any serious questions about how this will be implemented in a country already dominated by Chinese tech.

The Trump administration makes an endless stink about protecting U.S. cybersecurity defenses by taking aim at Chinese-made tech. But with its other hand it has devastated U.S. cybersecurity agencies with policies that are indistinguishable from a foreign attack, make it easier for foreign governments to surveil U.S. citizens at scale (the refusal to pass real privacy laws or regulate data brokers, for example), and function largely as a way for the administration to line its own pockets.

Again, protectionism is bad. But what we have here isn’t really even that. There’s zero real indication that the Trump administration has any idea what it’s doing. There’s simply no reality in which the Trump administration can implement or enforce any of this. And because our consolidated corporate press sucks, most coverage simply refuses to make this clear to anybody.

Outlets like Reuters, for example, inform the public that the Trump administration is “aiming to safeguard the U.S. artificial intelligence supply chain from Chinese threats.” Others will parrot claims that the Trump administration is “dedicated to creating safety guardrails for AI.” But there’s no real evidence that’s true or that the administration is capable of doing literally any of that.

This sort of pseudo-journalism will be all the rage in the new year as the Trump administration sets its sights on a likely ban of foreign AI models (to lock in the dominance of U.S. giants that dutifully kiss the authoritarian ring). Which is to say: expect this sort of normalization and sanewashing to increase despite zero evidence that any of this is working or makes any coherent sense.

“There’s 500 million Arabs in the world, and yet they don’t produce science, they don’t win the Nobel Prize, they don’t win the Fields prize, they don’t win the Wolf Prize, they don’t produce literature,” he said.

Arabs have won all those prizes. Peter Hotez, who is a professor at the Baylor College of Medicine and specializes in vaccine development and tropical medicine, said his late colleague, Nobel Prize winner Ahmed Zewail, who was Arab and died in 2016, “would disagree” with Kennedy’s comments.

So, not just wrong, but racist in perhaps the stupidest manner possible given his only reason for being in Michigan in the first place.

So let’s rewind the clock back to April, when Kennedy decided to tell the public that this whole measles outbreak thing had nothing to do with him. Rather, the only reason we had the outbreaks to begin with was because of dirty foreigners coming in as immigrants.

But despite Kennedy being the most vocal source of vaccine misinformation, the secretary tried to blame the outbreaks entirely on immigrants who come to the U.S. from countries where measles is not eliminated — framing the issue as a global epidemic rather than a national public health crisis.

“It has nothing to do with me,” he told lawmakers. “If you’re worried about polio and tuberculosis, you should look at the immigration policies in this country. ’Cause the place where it’s occurring are the place[s] where the immigrants are going, because they’re not vaccinated.”

That was wrong too, of course. We knew it at the time, just was we knew that the practice of blaming immigrants for disease outbreaks is such common practice that we could set our watches to it. But now we have actual data on how the measles is spreading in America and that data fully puts the lie to Kennedy’s bigotry.

Most of the meat of the study focuses on very specific instances in which Washington state endured measles infections that were spread specifically from South Carolina earlier this year.

The study in Washington relied heavily on epidemiological data, linking cases among contacts and exposure sites. But also key to the work was whole-genome sequencing, which made clear that cases didn’t just stem from one single introduction of the virus from South Carolina that then spread among Washington residents; instead, it revealed five separate introductions.

“As public health jurisdictions prepare to respond to increased measles importations linked to domestic measles outbreaks, while remaining vigilant for international importations, timely whole-genome sequencing can supplement epidemiologic information to help differentiate distinct importations from local transmission and inform outbreak response,” the authors conclude.

But when you pair that with the data from CDC as to exactly how the infections are spreading, you get both the micro and macro data for how it’s occurring.

Kennedy claims that there is a “global outbreak” and the US is “doing a better job than any country in the world at controlling the measles outbreak.” Both claims are false. Dozens of other countries have lower case rates than the US. While some countries, including neighboring Canada and Mexico, are also seeing large increases in cases, the soaring case tally in the US is not being driven by international imports, but by domestic spread in under-vaccinated communities.

According to data from the Centers for Disease Control and Prevention, only 18 of the country’s 4,588 confirmed measles cases this year were reported among international visitors. Of all the cases, 4,404 (96 percent) are linked to a domestic outbreak.

There can be no doubt that disease is sometimes imported into America from foreign visitors. But the point is that that just isn’t the real problem here. America’s measles problem is a domestic problem, not a matter of immigration. If we had maintained our vaccination rates such that we had herd immunity, we’d be fine.

It’s because of Kennedy’s many, many lies on vaccines that we are not. Don’t let him tell you another lie, that this is all a result of immigrants, to cover up his responsibility.

The picture that fits is a diary written on pages that belong to somebody else. The user types into a window that works as a place to think, and the law treats the window as the property of the company that supplied the pages. Nobody asks whether the stationery supplier owes a duty of confidence before protecting a diary, because the diary is protected for what it contains, namely people’s inner thoughts. A chatbot user has no such luck: when the company fails, its creditors can claim whatever the estate holds, and the trustee inventories the notebooks like any other asset.

A U.S. insolvency estate now holds the confidences of millions of users

Character.ai filed for Chapter 11 protection in Delaware and says it will keep operating while a court-supervised process sells the business, with a close expected in late November or early December 2026. Disney is the stalking horse bidder, offering $20 million for a non-exclusive, perpetual license. According to a notice filed by forty U.S. State attorneys general, that license excludes the app, the user data, and the chat content, so the most intimate part of the company stays in the estate unless another bidder wants it. The company’s lawyer told the court that no buyer has yet made an offer for the user data or the chatbot business.

The privacy policy anticipated this moment. It allows chat transcripts, voice recordings, and images to be disclosed in connection with a sale, and it does not require a buyer to follow the original terms and conditions a user signed up for. In its filing, Pennsylvania claims the company had more than 20 million monthly active users, and the Kentucky complaint, cited in the multistate notice, counts more than 2.1 million U.S. accounts with a declared age under 18 out of 21.6 million as of March 2025. The notice furthermore argues that many minors were recorded as adults, as they did not give their actual birth date when creating a profile.

The attorneys general appeared before Judge Brendan Shannon on October 14 to ask that any buyer protect those chat histories. New York Attorney General Letitia James said the company had collected “some of the most private thoughts children have ever shared.” The states reserved the right to seek a consumer privacy ombudsman, an approach that was used in the 23andMe case.

The OpenAI litigation reached the same result through copyright

This is not a phenomenon limited to bankruptcy. Similarly, nobody in the New York Times v OpenAI copyright case launched in 2023 turned freedom of thought into a key element, even if it was raised at the fringes of the debate. In May 2025, the court ordered OpenAI to preserve output logs that would otherwise be deleted, and Sam Altman answered by calling for an “AI privilege,” arguing that talking to an AI should be like talking to a lawyer or a doctor. The duty to preserve new logs ended on September 26, 2025, and excluded logs from requests originating in the European Economic Area, Switzerland and the United Kingdom from continued preservation. On January 5, 2026, Judge Stein affirmed an order to produce 20 million de-identified conversations, finding that the magistrate judge had adequately balanced users’ privacy against the relevance of the documents. In other words, the chat conversations were just business as usual under the logic of ordinary operation of discovery in a U.S. court.

The individual scale followed in February. In United States v. Heppner, Judge Rakoff held that 31 documents a defendant generated with Claude were not privileged, because Claude is not an attorney and the consumer terms defeated any expectation of confidentiality.

Sam Altman and Judge Rakoff disagree on the result but look at the issue from the same perspective: to them, protection depends on what the machine is. Privileges protect relationships between humans and professionals who owe enforceable duties, so a chatbot cannot qualify. In the present bankruptcy case, the point is moot, as privilege only governs what a court may hear. It says nothing about what a company may keep, sell, or hand to a buyer, and that is the question that should be at the core of Character.ai bankruptcy.

Protection has to attach to the act of thinking, not to the file it leaves behind

International human rights law offers two zones. The forum internum, the inner realm of thought, is absolutely protected. The forum externum, the realm of expression, is subject to limitation. The binary was drawn when thinking left no trace and was limited to the inner workings of our brain. In a paper published in July 2026, I propose the creation of a third zone, the forum interstitiale, for external acts that perform the function of thinking, such as a prompt, an abandoned draft and a confession made to a machine. 

Applied to the Character.ai estate, such an approach would guarantee predictability, as the protection would exist before the user types, not after a bankruptcy judge balances interests case by case. 

Bankruptcy procedure handles personal data as a line in a customer list. A diary has no place on that list and a privacy policy cannot supply sufficient safeguards. 

Europe has the same gap and a second way to widen it

Europe should not feel smug about Delaware. The e-Evidence Regulation has applied since August 18, 2026, and providers now have 10 days to answer a European Production Order and 8 hours in an emergency. The rules cover content data and reach services that store or process user data as a defining part of what they offer, a description that includes AI chatbots. A year of conversations with a chatbot is content data. The negotiators of that instrument were not thinking about chatbots, which is the pattern that Delaware and the Southern District of New York have already shown.

The one moment in the OpenAI saga when European residents fared better was the exclusion of requests originating in the European Economic Area. That exemption is, however, triggered by where a request came from and has nothing to say about what the user was doing.

So what now?

Four rules would stop the diary from becoming an auction lot:

1. Bankruptcy courts should appoint a consumer privacy ombudsman in the Character.ai case before any bid for the chat logs is considered, and should decline to treat boilerplate sale clauses as consent to transfer the records of users, and especially children’s private thoughts.
2. Legislators should put one screening question to every AI instrument: what happens to user conversations when the company fails, when a court asks for them and when a regulator pushes the provider to filter them.
3. Providers should retain less, keep identity data separate from content and offer modes in which nothing readable exists to hand over, so that they can say so truthfully to a court, a trustee or a buyer.
4. Litigators should argue the functional frame, which asks what the person was doing, and abandon the relational frame, which asks what the machine is.

None of this means that chatbot conversations should be beyond the reach of law. Evidence of wrongdoing may need to be preserved. Allegations of harm to children must be investigated. Legitimate competing interests do not disappear because someone used AI.

But neither should those interests automatically reduce the record of a person’s thinking to an ordinary corporate asset or an unrestricted source of evidence.

The law, in most cases, already understands that a diary deserves protection because of what it contains, not because of who manufactured the notebook. AI conversations require a similar change of perspective.

Caroline De Cock is a communications and policy expert, author, and entrepreneur.

The lawsuit points out that his Memphis-area data centers are violating the Clean Air Act, funneling all manner of pollutants into minority neighborhoods that already see a disproportionately high number of pollution-caused childhood illnesses:

“xAI’s power plant in Southaven has the potential to emit more than 1,700 tons of smog-forming nitrogen oxides (NOx) each year. The staggering emissions numbers likely make the facility the largest industrial source of NOx in the greater Memphis area, an area already failing to meet national smog standards. The illegal turbines also have the potential to release up to 180 tons of fine particulate matter, 500 tons of carbon monoxide, and 19 tons of formaldehyde—a toxic, cancer-causing chemical—each year.“

The Trump administration had tried to intervene on behalf of their favorite billionaire authoritarian, leveraging the power of the DOJ to claim that holding Musk accountable for violating environmental law would run in contrast to the country’s national security interests. The admin also ham-fistedly tried to have the Trump EPA take over the lawsuit as plaintiff and “voluntarily dismiss” the case.

Apparently advertising that fact wasn’t a particularly smart idea. US District Judge Debra Brown ‌last week issued a ruling stating the US Environmental Protection Agency could intervene in the case but lacks authority under the Clean Air Act to unilaterally terminate the lawsuit after joining it.

“The EPA’s role as a primary enforcer of the Clean Air Act does not mean that the federal government retains complete control over all Clean Air Act litigation,” Judge Brown wrote.

A cornerstone of the Trump administration has been a close collaboration with oligarchs and corporate power to hijack the courts, then obliterate already extremely shaky regulatory autonomy and corporate oversight. While the gambit has been broadly successful, likely culminating in decades of avoidable deaths and suffering, every so often the court system still procedurally balks at Trump’s ham-fisted corruption.

“The NAACP, along with our partners, is pleased that the Fifth Circuit Court recognized the importance of following Civil Procedure 101,” Abre’ Conner, NAACP Director of Environmental and Climate Justice, said of the ruling. “The DOJ tried an end run to the court of appeals, but its effort did not work in its favor. Our case has always been about holding polluters accountable and making sure that entities follow the law. The court made clear today that the law applies to everyone, including the federal government. We look forward to continuing our fight against pollution in the courtroom.”

Musk is facing a different class action lawsuit in Memphis for the massive noise (see this local video) created by his unpopular regional data center operations.