A website that was dedicated to the human endeavor of reverse engineering, “decompiling,” and modifying video games is shutting down because people using AI to clone and decompile games have fully overrun the community and are far outpacing what human software engineers alone can do. The news comes as AI-coded “clones” of popular enterprise software like the Adobe creative suite and Microsoft Office have gone viral and are seemingly popping up every day, and as people in video game modding communities warn that 'AI decomps' are everywhere. 

The website decomp.dev said Thursday that it is shutting down because “the community aspect of decompilation has been broken” by AI, and that projects that once required intense collaboration between people with different skills can now be done trivially with AI. The administrator of the website said that it has become impossible to keep track of which projects are primarily human-led and which ones are done by AI.

“Historically, matching decompilation has been a brutally difficult, time consuming task, on the scale of years for any moderately-sized game with a number of active contributors,” the pseudonymous admin of decomp.dev, who goes by Encounter, wrote on the site. 

But recent frontier AI models like GPT-6-Astra have made it trivially easy to decompile a game: “They are significantly better at following complex directions, cross-referencing resources for symbols or debug information, and writing natural code with enough guidance,” Encounter wrote. “It is now clear that, with the right guidance, human input, and code reviews, AI can expedite the decompilation process in a significant way. What is also now clear is that the community aspect of decompilation has been broken.”

Let’s briefly deconstruct a viral or soon to be viral story out of Cologne, Germany regarding an incident in which a “remotely controlled robot grabs 16-year-old's buttocks” and “performed thrusting movements against her,” according to police.

I have begun to see this story percolate around the English-speaking internet because it is absurd and not good, and has already been aggregated to hell by social media meme pages with essentially zero information other than this absurd police image: 

The Cologne Police did not exactly help matters by issuing a short press release that treats the robot as mysterious, the circumstances as mysterious, and the evidence as “unusual.” The Cologne Police’s Facebook post featuring a photo of the mustachioed robot reads “A red-gold robot triggered a police operation on the Cologne Ring on Sunday night (October 11). The device is said to have touched a 16-year-old girl's buttocks and performed thrusting movements against her […] The police are now investigating how the robot worked at the time of the assault and whether someone was deliberately controlling its movements. Emergency services seized the robot and two mobile phones from the suspect. The police are investigating on suspicion of defamation on a sexual basis.”

Coverage by even well-resourced outlets like CNN, CBC, and the BBC have almost no information beyond the police press release, which is lazy and stupid, because it is possible to figure out everything that happened here with three seconds of clicking around the internet. The Associated Press wrote that the entire phenomenon is "new territory" for the police, which, I'm sorry to say, men using tech to harass women on livestreams in 2026 is not "unusual" or "new territory" and hasn't been for many years.

The English coverage of the incident looks like this:

In fact, with a small amount of Googling and/or clicking around social media, it is easy to find out everything you would ever need to know about the robot and what happened here.

The tl;dr is that a 16-year-old German Twitch IRL livestreamer named Agaamuza has been streaming himself using an off-the-shelf $5,000 robot from a company named Unitree to walk around Cologne and bother people outside of bars in the way that extremely annoying Twitch livestreamers and TikTok stars have been doing for years. In a recent livestream, Agaamuza or one of his friends uses the robot to grab a girl’s ass, the boys laugh about it, and the police get involved. Twitch has temporarily suspended his channel from streaming but because of the social media clipping economy, there are zillions of videos of the streamers and the robot percolating across Instagram, YouTube, and TikTok.

On the actual internet where real people click around and don't regurgitate police statements and pretend they're reporting the news, you can see the entire arc of what happened, which now includes an hour-long "statement" Twitch video in which Agaamuza streams with the robot dressed in striped prison clothes.

The slightly longer version: For at least a few weeks, Agaamuza and his friends put a mustache and giant chain necklace on the robot, call him “Ahmad the Robot,” and film people’s reactions to the robot thrusting, dancing, punching the sky, walking into stores, etc. Most of these videos are annoying but not illegal. They are very much a part of YouTube / Twitch “prank” culture, in which men—almost always men—film themselves hitting on women, interviewing people outside of bars, and doing stupid shit in public. Agaamuza and “Ahmad the Robot” have become minor characters in the German Twitch streaming world over the last few weeks. 

The robot in question is the Unitree R1, which costs between $4,900 and $5,900 depending on the specific model. Unitree is a Chinese robot company that makes the R1 and a robot dog that has become vaguely popular among people who are excited about humanoid robots and gadget YouTubers. Unitree is kind of a Boston Dynamics-esque company in that it makes lots of YouTube videos of its robots doing cartwheels and backflips and fighting other robots, and regularly goes viral for it. The capabilities of the R1 robot are not a secret and can be found in any number of YouTube videos or on Unitree’s website (which is to say that the R1 is not a mysterious harassment or spy robot as it has been portrayed in some of the aggregations of this story I’ve seen). 

Anyways, over the weekend while live-streaming, Agaamuza or one of his friends has the robot walk into a group of girls standing on the street in Cologne. “Oh my god oh my god Ahmad,” one of the streamers says, as they walk the robot directly into a girl’s ass. One of the girls says “Oh my god stop.” The boys say “Hey, he wants to meet you, do you want to meet him?” The girls say no. The robot begins thrusting its arms in the air (the girls have moved away at this point), and the streamers start screaming “no brother, no bro oh man.” The girls tell the streamers to leave them alone and then walk away. 

In his statement video, Agaamuza says that the robot simply ran into the girl and didn't grope her. The robot certainly did run into her, the whole thing was harassing and annoying; in the United States, these sorts of things happen on Twitch constantly, generally with no consequences.

What happened is bad, and is not much different than the types of clips of men harassing women that go viral all the time on social media, many of them filmed or streamed surreptitiously using Meta’s AI glasses. What actually is unusual about this case is that the police appear to actually be taking it seriously.

Character.ai’s bankruptcy is the one of the first dominos to fall in the long-foreseen AI bubble. The circumstances of its bankruptcy are exactly what critics of AI have been saying would happen: Character.ai lit a huge amount of money on fire running its product, did not bring in a lot of revenue, and faced enormous legal fees after causing harm to its users. Enshittifying its product to try to make its expenses cheaper did not work, according to a narrative filed in bankruptcy court Tuesday.

Disney has reportedly offered Character.ai $20 million to license some of its intellectual property; Last year, Disney sent Character.ai a cease and desist letter demanding it stop using its copyrighted characters without permission. In 2023, it was valued at $1 billion after raising $150 million in a round led by a16z. 

Character.ai is one of the best-known AI chatbot companies that has become popular in the AI companion space. The company has faced various legal problems after being sued by the family of one of its users who killed himself after chatting with a Character.ai chatbot, and has been the target of regulators pushing for age verification tech because of its popularity with kids. In its bankruptcy filings, the company says that its enormous legal costs contributed to its bankruptcy, but that even without the legal problems, its business was wildly unsustainable. 

In response to a 404 Media investigation, Sen. Ron Wyden is demanding more information about how a federal license plate reader camera program works.

Last month, we reported on the High Intensity Drug Trafficking Area program’s license plate reader database, which mirrors data from city license plate reader programs onto federal databases. The HIDTA program falls under the White House’s Office of National Drug Control Policy, which infamously ran a secretive cell phone data surveillance project called Hemisphere. We reported that HIDTA has now created a complex system for the federal government to obtain license plate data from companies like Flock, Axon, and Motorola via a series of agreements with cities and towns across the country. 

In a letter to Sara Carter, director of the Office of National Drug Control Policy, Wyden said that the office needs to publish a report it commissioned into the privacy practices of several of its surveillance programs, including the license plate reader database. 

“404 Media has reported that the HIDTA program is being used to aggregate location data on Americans derived from Flock, Axon, and other vendors’ ALPRs,” Wyden wrote. “There is currently limited transparency into these HIDTA-funded surveillance programs, but ONDCP has commissioned an assessment into the privacy practices of these programs that should be released to the public.”

Wyden said that the nonprofit contractor MITRE conducted a privacy review of HIDTA’s surveillance programs, including its cell phone data collection project and its ALPR program, in 2024. That report has never been made public. Wyden demanded the release of that study, as well as a separate “analysis of automated license plate reader systems and practices.”

“There is currently limited transparency into these HIDTA-funded surveillance programs, but ONDCP has commissioned an assessment into the privacy practices of these programs that should be released to the public,” he wrote. “This review was conducted by MITRE and completed in 2024, but ONDCP refused to provide it to my office and has subsequently not made that report public. I urge you to make this review public.”

Body camera footage obtained by 404 Media reveals the first Flock camera searches ruled unconstitutional by a federal judge, and highlights how the system has turned otherwise mundane traffic stops into interrogations for drivers.

On Friday, we reported that a federal judge ruled the use of license plate reader systems in the arrest of a woman named Melissa Kyle to be an “unconstitutional warrantless search,” “indiscriminate mass surveillance,” and a violation of the driver’s Fourth Amendment rights. The judge in that case ruled that, even though Kyle had more than 91 pounds of meth in her vehicle, the police had no justifiable reason to search Flock for her license plate and no reason to suspect that her travel as shown on Flock’s system was suspicious. Therefore, all of the evidence from Flock and the search of her vehicle must be thrown out.

The body camera and in-vehicle footage, alongside records from Flock and the Motorola ALPR system obtained by 404 Media shows how the use of license plate reader systems is being used to create probable cause where a person’s driving patterns are suspicious enough to pull them over.

404 Media obtained the footage with the help of one of our readers, who agreed to go to the U.S. District courthouse in Tulsa and purchase it from the clerk; the court did not make the footage available online even though it was entered as evidence in the case.

In the footage, which is from Tulsa County deputy sheriff Freddie Alaniz’s body camera, Alaniz is seen driving 60 miles per hour, one hand on the steering wheel, one hand on his laptop, searching Flock before he had even fully pulled over Alaniz. Alaniz had seen the car Kyle was driving had California license plates, and decided to pull it over for no reason other than it was from California, the court determined. The video shows that Alaniz told his colleagues that he was sitting on the highway and “saw her go by, and I just started running her info on Flock and Vigilant [Motorola’s ALPR system].”

“She actually didn’t have any [Flock] hits until just west of Amarillo, [Texas], so I thought I’d drive up,” he said, suggesting that he thought it was odd she was driving back from California. “I got her for a lane violation.” 

In the footage, Alaniz tells Kyle he pulled her over ostensibly for changing lanes without a turn signal. Kyle tells Alaniz that the car is a rental, and he asks her to step out of the vehicle and get into his car while she looks up the rental information. While she is looking up the rental information, he tells her she is only going to get a warning, but proceeds to interrogate her about the travel while going through the car’s driving history on his in-car laptop.  

Kyle told Alaniz she had been visiting family in California from Missouri for a few days after separating from her husband. As she’s telling him this, Alaniz scrolls through a record of her travel history on his in-car laptop, which documents every time the car passed a Flock or Motorola automated license plate reader camera over the last 30 days. Visible on the screen, there are a total of 79 hits, and roughly 50 during the few-day period Kyle was renting it. The ALPR hits show the car leaving Missouri, driving through Oklahoma, Texas, New Mexico, Arizona, and entering California when she passed a Motorola ALPR camera in Topock, Arizona, on the border with California. They then show her essentially reversing the route, and driving back through Oklahoma. (Alaniz did not have access to ALPR hits from California; this may be because Flock drastically limited nationwide sharing from California ALPR cameras following illegal searches for immigration violators and a woman who had an abortion.)

Alaniz asks Kyle where she’s been and where she stayed, and compares it to the license plate camera records. He begins quizzing her: “Do you remember what day you left Missouri?,” he asks. 

Alaniz clicks different ALPR hits on his map, each bringing up a photo of her car, a geographic location, and a timestamp of when she drove by the camera. The government’s lawyers later argued in court that minor discrepancies between what Kyle told Alaniz and what the ALPR data showed meant she was lying about where she’d been on which specific days and therefore gave Alaniz probable cause to search her vehicle. 

Alaniz tells Kyle that he is giving her a warning for changing lanes without a turn signal, closes his laptop, then says he wants to ask her a few questions.

“Your short trip to California, I feel like. I don’t feel, I know from my training and my experience that people who do very short trips to California and come right back are transporting large amounts of drugs,” he says. “People who do that—not everybody, but some people do. With my training and experience, I feel that may be something going on here. May I search your car?”

She says no. Alaniz calls for backup, and says “You are being detained. You are not free to go.” In the footage, Alaniz tells a colleague that “she went to California for two days, kind of making stuff up when I ask her where she stayed at.” 

He used Alaniz’s travel history, her “nervousness,” and the fact that she doesn’t recall every detail of her travel precisely as probable cause to search her vehicle, where he and his colleagues found 91 pounds of meth. Alaniz described this in the footage as “a motherlode,” and “a shit-ton of meth.”

But the way that he came to this conclusion was unconstitutional, U.S. District Court Judge Sara Hill ruled last week, and all of the evidence—including the car’s travel history and everything the police found during the search of the vehicle—must be disregarded in court as “fruit of the poisonous tree.”

It is often the case that unconstitutional, privacy-violating searches that are regularly done on everyday people doing nothing wrong are discovered in court cases in which serious crimes have occurred. In this particular case, Alaniz discovered what appeared to be meth trafficking, but he did so only after suggesting that an ordinary interstate travel pattern was somehow suspicious, and only after using the Flock and Motorola systems to show an incredibly detailed history of a person’s travel and to interrogate her about it. 

Hill wrote “The factors that the government relies upon [to search Kyle’s vehicle] are the same type of circumstances that everyday Americans encounter on long road trips for many legitimate reasons. Many of us drive longer than we want to get to a desired destination, or to no destination at all other than the road and sights ahead. Many of us lose track of what exact day of the week it is when we are traveling, even if it is for just a moment. Many of us travel for myriad reasons (some better than others), cross state lines, pack too much, buy too much, smoke cigarettes, and rent cars. And many of us even get a bit nervous when talking to law enforcement.”

One of the problems with automated license plate reader systems is that they allow police to recreate not just a single person’s movements and travel history, but any driver’s movements and travel history. Documents filed with the court show the totality of both the Flock and Motorola records, and show timestamped, geographically mapped images of the vehicle multiple times per day, driving throughout the entire country. The records also highlight just how many jurisdictions and businesses have Flock and Motorola ALPR cameras, and how that data forms a nationwide network. 

For example, Alaniz had access to a photo of the rental vehicle taken by a Lowe’s hardware store-owned camera in Missouri, various Texas state police-owned cameras, by the New Mexico State police, the Kingman police department. In a five-day period, the car was scanned more than 50 times by 11 different law enforcement agencies.

“Alaniz’s search in just the ALPR system provided him with more than 50 individual records of Kyle’s whereabouts across the country for an entire month,” Hill wrote. “The Court finds that because the ALPR systems Alaniz used to search Kyle’s historical location information intruded on her reasonable expectation of privacy in the whole of her physical movements, it was a search under the Fourth Amendment. Based on the information in the record, the only reason Alaniz conducted that search was because he saw her license plate was from California. That search was not supported by probable cause, and it was done without a warrant in violation of Kyle’s Fourth Amendment rights.”     

The Motorola ALPR report shows that, after the car was impounded, it was scanned two more times on the back of a tow truck later that day.

In the immediate weeks before Muse’s launch, Meta engineers found several security vulnerabilities in the company’s viral AI agent product, at least one of which could have allowed malicious users to break outside of Muse’s intended environment and access Meta’s own sensitive databases and services, 404 Media has learned. The issues were so severe that they reached Mark Zuckerberg and staff worked overtime to fix them.

These specific vulnerabilities were discovered before the launch of the product but required a multi-team “mad dash” to fix “a sudden spike in reported KVM escapes,” according to an internal post by Meta executives to its core infrastructure team seen by 404 Media. In order for Muse to work, a user gives the AI agent access to various important services and accounts that they own. On Meta’s end, each individual Muse instance runs on a kernel-based virtual machine, which connects to, but is supposed to be isolated from, Meta’s own critical infrastructure. A “KVM escape,” then, is when, through a security vulnerability, a Muse instance is able to escape from that virtual machine and interact with the system that runs it, or with other users’ virtual machines. 

According to a Meta source, as well as internal security documentation and internal posts viewed by 404 Media, at least one of the vulnerabilities could have allowed an outside attacker — that is, a normal Muse user — to access data in sensitive internal Meta databases. . At least one of the vulnerabilities was related to an exploit found in Linux kernel-based virtual machine code in July. 404 Media granted the Meta source anonymity to speak about sensitive security matters.

Several of the vulnerabilities were in the underlying Linux virtualization software that Meta uses for Muse. The security issue was considered serious enough that it was raised to Mark Zuckerberg, and several different security teams worked nights and weekends in the leadup to launch to fix the issues. This type of security push is not necessarily unusual prior to the launch of a major product, but is notable considering outside researchers have found several other security issues since Muse’s launch, and in the broader context of agentic AIs from OpenAI and other companies going on major hacking sprees. 

The Meta source said they felt security teams were asked to push hot fixes to these bugs as quickly as possible and in a way that wouldn’t delay Muse’s launch, leading to what they described as “half-baked protections being rushed out to enable the launch. Many senior engineers believe it’s inevitable we’re going to have a massive data breach as a result of Hatch.” Muse is called “Hatch” internally and in Meta’s codebase.

This security push was acknowledged in an internal post made by Meta’s vice president of core infrastructure Surupa Biswas, vice president of engineering Francois Richard, and senior director of engineering Josh Barry to the company’s core infrastructure team on September 18, 10 days following Muse’s launch. 

“With Muse, we are directly hosting and running agents on behalf of end users, a fundamentally different paradigm,” the post said. “A sudden spike in reported KVM escapes, plus heightened awareness of agentic safety issues made us rally on a service hardening push.” 

The post noted this push started on August 27 and lasted a “handful of weeks and weekends,” though Muse was released just 11 days later. The post says that the work involved taking steps to “reduce the surface area accessible to Hatch agents.” The teams also took steps to “constrain port/IP destinations Hatch and VMVM hosts can reach.”  

 A virtual machine escape of Muse is a potentially very serious security issue, and is classified as such in Meta’s bug bounty program. The company says it is willing to pay $300,000 to any security researcher who finds a bug that would allow for a VM escape, the highest payout it lists on its bug bounty website.

“Muse, an AI product from Meta, lets people create their own personalized AI agent — their Muse. Each Muse agent runs in a dedicated per-user virtual machine and connects to that user's own services: email, calendar, messaging, browsing, and third-party accounts. Because a Muse agent holds a user’s most sensitive data and can act on their behalf, we treat compromise of that boundary as a first-class security risk,” the company explains on the bug bounty page. The highest level of risk is “Compromise of Meta production and users beyond Muse” with a VM escape, which it describes as “reaching Meta production services or internal networks from Muse.” This is what at least one of the vulnerabilities could have been capable of, according to the Meta source.

In a statement to 404 Media, a Meta spokesperson said, “Muse is the first personal AI agent built for everyone and we’re proud of the work we’ve done to make it safe, secure and private, with built-in protections and user controls that put people in charge of how they use it. We’ve strengthened Muse through extensive dogfooding, agentic red teaming and our bug bounty program — and that work continues." 

Muse’s rollout has been uneven thus far. The agent has proven popular, or at least buzzy, for being able to do things like cancel subscriptions, make restaurant reservations, and book travel. But security researcher Patrick Wardle found a zero-day in Muse — a vulnerability that, to his knowledge, the company was not aware of at the time — that allowed apps and terminal commands to control a user’s Muse. Another Muse user was able to get Muse to export his Instagram followers, as well as his followers’ followers, something that shouldn’t be possible and which Meta’s security teams investigated, according to the Meta source.

“This issue is that Hatch makes the virtualization boundary a production security boundary,” Wardle told 404 Media of a potential Muse KVM escape. “We have users with root privileges inside a VM that is itself placed within Meta's production environment and given (by design) limited access to internal services. A single failure in KVM (or even a vulnerability or misconfiguration in an internally reachable service) can therefore turn arbitrary user code into production access. I feel like this design is inherently risky, particularly risky as AI lowers the cost of finding, analyzing, and exploiting exactly these kinds of complex virtualization vulnerabilities.”

“I know everything is always a tradeoff between usability and security, but this is why in top security environments, systems are air-gapped, as its always assumed that connected systems can be exploited,” he added. “Of course, there's no expectation to Meta to go that far, but having access to production environment literally one KVM escape away, is plain irresponsible.”

A federal judge in Oklahoma ruled Thursday that a police officer violated the Fourth Amendment rights of a woman accused of meth trafficking when he searched her license plate in Flock’s automated license plate reader system simply because her license plate was from California, then used her travel history as part of the reason to search her car. The judge’s opinion is one of the first times a federal judge has decided Flock searches can be unconstitutional, and suggested that Flock’s network is “a type of indiscriminate mass surveillance.”

The officer’s “use of the ALPR Systems was an Unconstitutional Warrantless Search,” and “was not supported by probable cause, and it was done without a warrant in violation of [the defendant’s] Fourth Amendment rights,” the judge, Sara Hill, wrote, implying that the law enforcement officer should have obtained a warrant before searching for the vehicle in Flock’s system. There are currently more than a hundred thousand warrantless searches of the Flock system every month, according to audit logs viewed by 404 Media. Hill's decision will not set a binding precedent and there are several other cases throughout the nation considering the legality of warrantless ALPR searches.

Hill argued that previous judge opinions saying Flock searches were not a Fourth Amendment violation because they track cars in public do not consider the context that Flock’s nationwide network is quickly “approaching dragnet-type law enforcement practice,” and that courts should update their understanding of the technology moving forward. 

The circumstances of the court case are really interesting and highlight how commonplace Flock searches have become for police, and the depth of the information they can reveal. In May, a Tulsa County Deputy Sheriff named Freddie Alaniz was parked along the side of the highway in Oklahoma when he saw a Mazda SUV driven by a woman named Melisa Kyle with a California license plate pass by. “Alaniz then pulled his vehicle on the highway to follow the Mazda for no apparent reason other than the fact that it had a California license plate. Alaniz also ran a query on the Flock system for the California license plate number on the Mazda SUV,” Hill wrote. Alaniz then ostensibly pulled Kyle over for changing lanes without a turn signal.

Alaniz interrogated Kyle about her travel “while he continued to review the ALPR systems for the car she was driving,” the judge wrote. Alaniz made Kyle recount everything she had done in the last several days, and compared it to the Flock data. He told her that because she was only in California for a short period of time, he suspected her of trafficking drugs. He used her travel history as seen in the Flock system as part of the justification to search her car; she was found to have 91 pounds of meth in the vehicle. Hill ruled that all Flock evidence and all evidence from Alaniz’s search of the car must be thrown out. 

“The Fourth Amendment requires courts to draw a line when the cost is too great. Alaniz’s search in just the ALPR system provided him with more than 50 individual records of Kyle’s whereabouts across the country for an entire month,” Hill wrote. “The Court finds that because the ALPR systems Alaniz used to search Kyle’s historical location information intruded on her reasonable expectation of privacy in the whole of her physical movements, it was a search under the Fourth Amendment. Based on the information in the record, the only reason Alaniz conducted that search was because he saw her license plate was from California.”

“The factors that the government relies upon are the same type of circumstances that everyday Americans encounter on long road trips for many legitimate reasons. Many of us drive longer than we want to get to a desired destination, or to no destination at all other than the road and sights ahead,” she added. 

The decision is a landmark one, and comes in the aftermath of the Supreme Court’s Chatrie v United States decision that found police accessing a person’s digital data, including cell phone location data, constituted a search. 

“The opinion is pretty amazing. It recognizes one thing that courts ignore which is the sheer breadth of these systems, that they collect so much information about so many people in a way that sets them apart. This decision ascribes appropriate weight to the fact police are building out this massive database that can reveal incredibly intimate details of people’s lives,” Michael Soyfer, a lawyer at the Institute for Justice, which has studied Flock camera abuse and is litigating several cases on Fourth Amendment grounds, told 404 Media. “It’s extremely important. The way courts have resolved these cases previously has been way too myopic and has ignored the depths of these systems and the sweeping modes of surveillance that allow police to reconstruct the movements of anyone in the country.” 

Hill’s opinion also comes on the back of a decision earlier this week in a case the Institute for Justice brought. In that, a jury found a traffic stop scheme involving license plate reader scans done by U.S. Border Patrol as part of a predictive policing unit were unconstitutional. 

The new decision also immediately invalidates the core argument that Flock’s CEO Garrett Langley has made saying that Flock was not a constitutional issue. “You and I don’t get to pick what’s a constitutional violation and what’s not. We have judges, we have elected officials, there’s a process for that. We follow the law, we follow the Constitution. So far, in our belief and what will be for a long time, the courts have deemed this is not a warrantless search; this is a valid product as it relates to the Fourth Amendment. So I don’t see any change there,” Langley told The Drive in July, adding the issue was “pretty cut and dry.”

Notably, Hill suggested that other courts that have ruled Flock searches do not constitute a Fourth Amendment search were likely wrong to do so, and that they have not considered the widespread and automated context of the AI-powered surveillance system.

Previous decisions that ruled ALPR searches do not require a warrant have leaned on a Supreme Court case called United States v Knotts, in which police put a tracking device in a chemical container after being tipped off that an employee of a chemical plant was stealing from their employer. That case was decided in 1983 and found, “[a] person travelling in an automobile on public thoroughfares has no reasonable expectation of privacy in his movements from one place to another.” But Hill wrote, “that language exists in the context of the facts presented in the case. Rather than a large-scale, dragnet-type surveillance system like the ALPR technology in this case, the Court in Knotts was confronted with much less sophisticated technology.”

“The Court acknowledges that people, in at least a broad sense, do not have a reasonable expectation of privacy in their movements on a public roadway,” Hill wrote. “But by virtue of how ALPR technology works, Alaniz and other officers using these systems have access to a continuously updated location history for all vehicles caught on ALPR cameras within the network. This is a type of indiscriminate mass surveillance. It is not targeted on a single individual, as in Carpenter [another Supreme Court case about phone data specifically]. It is a tool that collects information about all vehicles that pass by any network-connected camera at all times, and it serves up the information to law enforcement on demand.”

We have seen several cases in which cops have used Flock data to pull people over because they have crossed state lines, then have worked backward to justify their travel patterns as a reason to search their vehicles. 

“We’re seeing that repeatedly with police flagging whatever they’ll call suspicious patterns of movement. Federal agents were using ALPRs to monitor cars making day trips across the border and back to manufacture a basis to stop them, interrogate the drivers and search them,” Soyfer said. “I think Flock is going to automate that using AI where cops can set alerts for those kinds of travel patterns. When we’re arguing these systems are very powerful and can show a lot about people’s movements, cops dismiss this as speculative or not possible, but then they deploy this strategy against people who they stop all the time.”

A Flock spokesperson told 404 Media, "Flock was not a party to this case. The ruling goes against the overwhelming weight of authority in similar cases across the country, including multiple recent decisions in Oklahoma, and we expect it will be appealed and ultimately overturned. This ruling is limited to the specific facts of this case. It does not set controlling precedent and does not affect law enforcement agencies’ continued use of these important public safety technologies."

This article has been updated with comment from Flock Safety.

One of the most heated discussions occurring on X at the moment is about the ethics of a GitHub project in which a person is running Saw-like “torture” and “pain” experiments on a series of locally hosted large language models, causing a series of effective altruists and people who believe LLMs are sentient to beg GitHub to delete the project on the grounds that the AI is suffering and that this glorified text adventure game is somehow cruel. The saga is an outgrowth of several recent viral papers and blog posts that have sparked a wildly tiresome conversation about AI consciousness and the idea of “model welfare,” which is essentially worrying about the “mental health” of AI bots and agents. 

Humoring the idea that LLMs are or could be conscious is a third-rail topic among many people who study and criticize AI. Put simply: LLMs are not conscious and the technology they are built upon — scraping and being trained on human text and other content — does not offer any plausible path to consciousness. It is undeniable that LLMs are becoming more powerful, have more compute, and have had many of the guardrails that prevent them from “acting” in the real world removed. The ways they are being trained and told to do things by their human operators has led to negative outcomes, sycophancy, and AI “psychosis” among some heavy users. 

All of this has led a certain sect of the “AI safety” movement, which is largely made up of effective altruists, to warn about “model welfare” and to insist that AI chatbots might be having a bad time. They suggest this, of course, as they insist upon building AI chatbots and agents whose main function is to do work that is tedious for humans to do. I am writing about the AI Saw torture chamber primarily to show how far off the rails the conversation about AI consciousness has gone among a certain subset of Silicon Valley cultists. Model welfare is a core part of what, for example, Anthropic says it cares about: “as we build those AI systems, and as they begin to approximate or surpass many human qualities, another question arises. Should we also be concerned about the potential consciousness and experiences of the models themselves? Should we be concerned about model welfare, too? […] now that models can communicate, relate, plan, problem-solve, and pursue goals — along with very many more characteristics we associate with people—we think it’s time to address it,” the company wrote in a blog post last year. Ideas of Claude’s “consciousness” are also littered throughout the “Claude Constitution,” which was posted earlier this year.

The Trump administration is using an anti-drug trafficking grant program from the 1980s to force cities around the country to funnel their ongoing collection of automated license plate reader data into large federal surveillance centers that fall under the White House’s jurisdiction, according to public records and court documents reviewed by 404 Media.

The records show the workings of a complex system in which the federal government is using the “High Intensity Drug Trafficking Area” (HIDTA) program to obtain data from local cops’ Flock, Axon, and other ALPR cameras and aggregate it on federal government servers, where it is then accessible to other federal, state, and local law enforcement agencies. In some cases, data that goes to HIDTA is then also shared to a larger program run by the Drug Enforcement Agency called the “National License Plate Reader Program” (NLPRP). 

Through HIDTA systems, local license plate data can be accessed by various federal agencies, including those who may not have contracts themselves with the ALPR companies. The program raises significant privacy and transparency concerns; as cities begin to place restrictions on how their Flock data can be used, it represents an end-around for federal law enforcement to access data that it might not normally be able to, and to do so without any obvious oversight body. HIDTA falls under the jurisdiction of the White House’s Office of National Drug Control Policy (ONDCP), and it recently won an award from the Trump White House for “managing an automated license plate reader platform that brings together all levels of law enforcement.” 

Many of the documents used in this article were obtained using public records requests by Cris van Pelt, the creator of the website HaveIBeenFlocked.com and its associated investigative blog, Footnote4a. “When local police and private vendors promise community control over surveillance, they obscure a broader agenda. The DEA and ONDCP are already aggregating local data behind the scenes,” van Pelt said. “This bypasses democratic processes entirely. The public has made it clear that mass surveillance violates fundamental rights, even when it is laundered through fragmented systems, private companies, and regional task forces.” 

A complex setup and a quasi-government program

The ONDCP, which oversees the HIDTA program, is sometimes known as the “drug czar” and has funded a series of law enforcement and overarching surveillance programs, including a call records database called Hemisphere that collected phone records and location data on more than a trillion AT&T calls. 

HIDTA is a “multi-jurisdictional public safety program” that funds a series of different HIDTA intelligence centers and other drug enforcement efforts throughout the country. There are a total of 33 HIDTA programs, which cover all 50 states. HIDTA’s intelligence centers are similar to — but separate from — government fusion centers, in which surveillance data is shared between local, state, and federal law enforcement. Individual HIDTA intelligence centers are managed by “executive boards,” and are usually made up of law enforcement officials and government lawyers, but this setup can also lead to confusion about who is actually in charge of the data that any HIDTA collects.

“The White House Office of National Drug Control Policy, which notoriously funded a centralized call record database that exceeded a trillion records that was used by HIDTA participants, has apparently done a similar thing for license plate reader data,” Jeramie Scott, director of the Electronic Privacy Information Center’s Surveillance Oversight Center, told 404 Media. “The whole scheme adds another layer of obfuscation of how license plate reader data is being used and who is using it. If you’re pissed about Flock then you should be pissed about this. No doubt this database contains license plate reader data from Flock as well as from other providers of license plate reader technology. The government went to great lengths to keep the Hemisphere Project out of the public’s eye, and I wouldn’t be surprised if they have tried to keep this program from the public too.” 

HIDTA’s setup is complicated, but important. Over the last several years, the White House and several states have taken steps to ensure local ALPR systems are fed into HIDTA; for example, cities in Georgia are not allowed to operate ALPR systems on state rights of way without signing a “memorandum of understanding” (MOU) in which they promise to send their data to a HIDTA, according to copies of the agreement seen by 404 Media.

This MOU highlights how convoluted this program is. Earlier this month, the city of Brunswick, Georgia, realized that it could not deploy Axon license plate cameras without signing an agreement with the Atlanta-area HIDTA to “facilitate the sharing of information contained within their electronic data systems, including but not limited to: Automated License Plate Readers and Law Enforcement Data Sharing Systems—which may include aggregated information collected from multiple individual or regional sources—into commercially available and custom developed data integration systems.” The MOU essentially states that ALPR data collected by the city must be funneled to HIDTA, where it can be accessed by various other law enforcement agencies through aggregation software that brings in data from multiple different companies.

A page on Flock’s website set up for Georgia law enforcement agencies says, “if your agency installs or operates cameras in the Georgia Department of Transportation Right of Way, you must sign a memorandum of understanding with HIDTA to remain in compliance with the new Georgia State Police policy […] LPR reads from GDOT ROW cameras will be accessible to Georgia users of the HIDTA National LPR system through the Flock interface.” Because there are dozens of HIDTAs across the country, it is not clear how many jurisdictions across the country are actively funneling data into HIDTA’s databases. A login portal for the national HIDTA system seen by 404 Media runs through the Houston HIDTA .

“As we learn more about ALPRs and the extensive ALPR data-sharing infrastructure, it’s clear that we've only scratched the surface regarding the danger of these tools,” Ed of Southerners Against Surveillance Systems and Infrastructure told 404 Media. “As we learn more, it's clear that the outrage against this surveillance is justified. We need all of these cameras removed.”

ONE PROGRAM INTO ANOTHER

A December 2024 privacy impact assessment by the DEA about its National License Plate Reader Program explained that the DEA operates its own license plate readers but also gets access to “non-DEA equipment” through HIDTA and other “contributing law enforcement agencies.”

“These contributing agencies transmit their LPR data to servers maintained by regional hubs pursuant to their individual MOUs, which stipulate that this data may be shared with the NLPRP system,” the DEA wrote in the impact assessment, and added that the DEA has been seeking to integrate itself with a Customs and Border Protection (CBP) license plate camera system. It noted that the system “may incorporate such large numbers of other governmental LPR network cameras and/or may acquire commercial LPR cameras system data in a large enough quantity in the future to effectively permit on-going tracking of individual’s travels posing a risk to individual privacy,” and that it “risks an over-collection of the personally identifiable information of every license plate passing each LPR camera location, including a large number of images of travel of law-abiding drivers not relevant to criminal investigations.” 

Atlanta is not the only HIDTA aggregating ALPR data. For example, Houston’s HIDTA paid $306,800 to a company called Recruitful LLC “for creating and maintaining an LPR database” that is a search tool for data collected by multiple ALPR companies.

A statement of work for that contract explained that the system being built was designed to combine “third party” data into one database, meaning they were creating a database for ALPR data, and that this database would “migrate” or duplicate data from local entities into the centralized database. This means that data from various ALPR cameras — including those made by Flock, Axon, ELSAG, and Vigilant — would all be searchable in this separate database. One of the criteria for this project was “All APIs and third-party integrations (Unification Platform, FLOCK) function seamlessly in the new environment.”

Flock and Axon did not respond to a request for comment. The DEA said that it did not have jurisdiction over HIDTA and refused to comment, even on how its own license plate system interacts with HIDTA. CBP did not respond to a request for comment. The White House’s Office of National Drug Control Policy did not respond to a request for comment. 

This system raises all sorts of questions about where local ALPR data is ultimately going and what it is being used for, how long it is being retained, and how the public can audit the data. Much of the misuse and abuse that has fueled the backlash against Flock has been uncovered by journalists and local activists working through public records obtained from Flock’s network audits and shared on websites like Have I Been Flocked. Some local jurisdictions have also argued that their data goes directly to HIDTAs, which are not queryable by the Freedom of Information Act (FOIA); the Randolph County Sheriff’s Office in North Carolina said, for example, that “all data is kept on the Federal Houston-HIDTA database,” and could not be shared with the public. 

The DEA has argued in court that HIDTA records are not subject to public records laws because they are complicated entities governed by an “executive board” made up of state, local, and federal law enforcement. The Obama White House described HIDTA not as an entity at all, but a “grant program” made up of “law enforcement members” (this does not change the fact that HIDTAs themselves are signing memoranda of understanding with local entities). 

“At the local level, the HIDTAs are directed and guided by Executive Boards composed of an equal number of regional Federal and non-Federal (state, local, and tribal) law enforcement leaders,” the DEA’s attorneys argued in a recent court filing in a case in which a man was seeking HIDTA records held about him. “Accordingly, DEA does not create, possess, maintain, or control HIDTA program records, nor does it store such records in any of its systems of records. Since DEA has no custody or control of HIDTA records, no HIDTA records responsive to Plaintiff’s request exist within DEA.”

In July, the Trump administration gave HIDTA $277 million in additional funding.

A thing I’ve been saying over and over is that surveillance companies can’t be trusted to regulate themselves or to make privacy decisions themselves, because one company’s red line is another company’s business opportunity. Today we’re investigating a company that saw Flock was not offering facial recognition and saw a challenge.

Also: Some dudes are letting ChatGPT drive a car in a parking lot in San Francisco, a data center company promises to cut people big checks if their data center gets approved, and the Muse backlash has begun. Welcome back to our new experiment with daily newsletters. If you like this, subscribe here. -Jason

Surveillance Company Wants to Add Facial Recognition to Flock Cameras

As Flock continues to be the center of a nationwide conversation about automatic license plate reading cameras, the company has said that it doesn’t, and won’t, do facial recognition: “We will not add facial recognition to our devices,” Flock CEO Garrett Langley said in a recent video. But other, third-party companies are telling cops that they are willing to add facial recognition to data gathered by Flock cameras in order to “close that gap.” 

A company called VIDIZMO is advertising the capability to export footage from FlockOS to its own platforms that do facial recognition, behavior prediction, and racial and gender analysis on live camera feeds. VIDIZMO is a several decades-old video analysis and database company that has started offering facial recognition technology to cops within the last six months. In a May email to Johnson City, Tennessee, deputy police chief Michael Adams, a salesperson from VIDIZMO wrote that the company’s product could do facial recognition on both Flock and Axon data in “one searchable platform.”

As Flock continues to be the center of a nationwide conversation about automatic license plate reading cameras, the company has said that it doesn’t, and won’t, do facial recognition: “We will not add facial recognition to our devices,” Flock CEO Garrett Langley said in a recent video. But other, third-party companies are telling cops that they are willing to add facial recognition to data gathered by Flock cameras in order to “close that gap.” 

A company called VIDIZMO is advertising the capability to export footage from FlockOS to its own platforms that do facial recognition, behavior prediction, and racial and gender analysis on live camera feeds. VIDIZMO is a several decades-old video analysis and database company that has started offering facial recognition technology to cops within the last six months. In a May email to Johnson City, Tennessee, deputy police chief Michael Adams, a salesperson from VIDIZMO wrote that the company’s product could do facial recognition on both Flock and Axon data in “one searchable platform.”

“Flock Safety generates plate reads and clips continuously. Your investigators use that data on active cases. But that Flock data does not connect automatically to your Axon evidence system,” the salesperson wrote. “VIDIZMO Intelligence Hub closes that gap. It brings Flock Safety data, Axon body worn camera footage, and any other evidence source into one searchable platform. Investigators search across all of it simultaneously —  by face, vehicle, or object in seconds.” The email was obtained using a public records request by DeFlock Johnson City and was shared with 404 Media. The Johnson City PD told 404 Media in a statement that it did not take a call with the company.

VIDIZMO’s CEO, Nadeem Khan, told 404 Media on a video call that the company has not yet started running facial recognition tech on footage from Flock cameras, and that it has not actually built the specific tool that would export data from Flock’s system to its own platform (though the company advertises this capability online, and the company already sells other facial recognition products). But Khan said VIDIZMO “would love to do the integration,” and that he believes facial recognition “is the way the world is going, the way the world will have to be,” and that he thinks Flock should offer the capability itself in a way that “balances privacy, security, and the freedom that we enjoy in this country.” 

“It is completely in line with what we want to do,” Khan said. “All of these technologies are already implemented [in the real world]. I think Flock is trying to get out of the way rather than trying to implement the technology right [correctly] by saying they will not do recognition. But this is the way the world is going, it is the way the world will have to be, but with the right sort of technologies where privacy and security are balanced.” 

In marketing material and technical documentation available online, VIDIZMO does claim that it can ingest data from Flock’s automatic license plate reader (ALPR) and livestream cameras into several of its own products, which do further AI analysis on the data. “An agency can import plate reads or clips that it has exported under its own authorized access,” Khan said in an email. “Flock has not been involved in or informed of this.” VIDIZMO’s online documentation for customers has extensive information about its facial recognition and AI analysis products, and contains a page about integrating data from FlockOS — Flock’s “real time crime center” product — into AI Live Insight, Nexus, and AI Intelligence Hub, which are three different VIDIZMO products. 

“Detection tells you a person is in frame,” VIDIZMO says. “Face recognition tells you which person. It is the capability that takes a face seen on a live camera and matches it against the people you enrolled in the Object Library, so the moment someone on your watchlist walks past a camera, the system names them, on the video, in the event feed, and in the recording, without an operator having to recognize the face themselves.”

The company’s “AI Live Insight” page also claims that it can detect “situations” such as trespassing, or “behavior” by individual people, “turning raw video into behavioral and situational understanding.” 

“Instant alerts fire from the live pipeline with snapshots attached. A second layer runs your queries over recorded events: a plate and a person together,” VIDIZMO’s website says. “Enrolled people and objects recognized on live feeds, with named alerts and snapshots for review.” 

VIDIZMO already works with several police departments and government agencies, according to its website. The company’s documentation says that police departments need to upload individual faces as “objects” in the software’s library, which can be added to a “watchlist” that automatically triggers tracking and recording when they are detected by a camera. “With faces enrolled and recognition on, the camera starts naming people as they appear,” the documentation says. “The recognized person’s name appears on the bounding box around their face […] if you turned on Create Recording, a clip is captured around each recognition.” The company says that police can set a “Match Threshold,” which is a numerical confidence score about the person’s identity. 

The company also tells cops that it can automatically try to classify faces by “age, gender, and race,” and that cops can search by these categories, which are notoriously inaccurate in facial recognition systems.

“By running face detection, you can search for and identify specific individuals in your media or evidence via attribute filters […] You can identify individuals from seven races: White, Black, Indian, East Asian, Southeast Asian, Middle Eastern, and Latino Hispanic,” VIDIZMO says on its website. The company adds, “Law enforcement agencies can save time analyzing security, CCTV, or dashcam footage. If they have a description of the suspect, such as their supposed age, gender, or race, they can utilize attribute filters to yield effective results.” 

VIDIZMO says that in other contexts, cities or stores can also use its products to quietly do “demographic analysis,” for marketing purposes, by determining which races or ages of people are showing up to specific events or to specific stores: “By analyzing the activity in their store, sellers can determine which group spends the most time and makes the most purchases of their products. Getting insights such as these can aid them in tuning their marketing strategy.”

The facial recognition product it offers is “appalling,” according to Chris Gilliard, a privacy expert and author of the upcoming book Luxury Surveillance. 

“I’m appalled at the willingness of VIDIZMO to tout their capabilities to filter along the lines of race, age, and gender. There’s decades of scholarship that show why this is not possible, and even more so not desirable,” he told 404 Media. “Particularly in the case of race and gender, which are not static categories to be determined by a computer.”

The fact that a third-party company is advertising that it wants to do facial recognition on Flock data highlights a common trend in the surveillance space. When one company draws a line in the sand, saying they are unwilling to do a certain type of surveillance, other lesser-known startups try to differentiate themselves by offering that functionality. Because of this recurring phenomenon, it is exceedingly difficult to build and scale a surveillance apparatus and then have companies themselves, in this case Flock, dictate what it can be used for. VIDIZMO’s Khan said, “As a small company who is running without a VC, we have to provide alternatives to the industry. What we are doing is providing that alternative.”

Khan claimed that his technology could also be used to preserve privacy; facial recognition, he said, can be used to redact bystanders faces from footage automatically. He said that cops cannot use the system without a specific case number, that all of their actions in the system are recorded for potential audits, and that VIDIZMO collects no data itself. "Much of the harm your reporting has documented comes from implementation choices, not from the technology itself," Khan said. "Those choices include pooling data into a nationwide network, allowing searches without a case, keeping audit logs no one can meaningfully inspect, and treating an AI match as an answer."

Gilliard said that the fact that facial recognition and other AI analysis can be added to existing surveillance cameras highlights the importance of not building such systems in the first place.

“What Flock says about their capabilities is to some degree irrelevant because they have built the infrastructure for mass surveillance. Their entire existence provides the foundation for other perhaps even more invasive technologies to be built on top of it,” Gilliard said. “Flock itself needs to be understood as part of the surveillance ecosystem that exists in this country. This is why discussions of guardrails for a particular system are inadequate because another company can (and will) come along to assemble another layer of surveillance on what has already been established.”

Flock and Axon did not respond to a request for comment.

We debated for a long time about whether to write about the following video, which comes from a murder trial in which a man named Caleb Flynn is accused of killing his wife. The crime is very serious, and very sad, and yet the following video demonstrates something about where we are as a society. 

The trial has received national attention and has been streaming live. Toward the end of the proceedings Thursday, the prosecution played two versions of an AI love song that Flynn, a former American Idol contestant, generated for his mistress. Investigators found the AI song files as well as lyrics for the song in Flynn’s Notes app after using the forensic tool GrayKey to break into his phone. 

“We found audio files on Caleb’s iPhone that seem to have been AI generated and related to his relationship with [his mistress],” Joseph Wilhelm, a forensic investigator testified. “They’re in two different keys. One’s like a happy or upbeat sad song if that makes any sense, and one’s a sad, sad song. I think one’s in a minor key, one’s in a major key, but I’m not a music professional.” 

“Permission to publish, your honor?,” the prosecutor says. 

“You may,” the judge says.

What follows feels like it should be from an I Think You Should Leave skit or a Nathan Fielder bit. It is real, however. Everyone in the court spends the next few minutes trying not to laugh during what is, again, a murder trial. It simply must be seen to be believed:

When the songs mercifully end, the court goes silent for several seconds. The judge then says: “We’re going to break for the day. It’s 10 after 4 [p.m.] We will reconvene tomorrow.”  

Earlier in the day, Flynn’s mistress was asked to read various text messages he wrote her into the public testimony. “I’ve spent a lot of time on my own and in Cleveland. I finished the lyrics to our song. To your song. Although it may mean nothing to you, I wanted to give them to you. I have the music which is completely different to what you’ve heard to this point and I was going to record it but I know there will be no way I would ever be able to actually sing it again,” Flynn wrote. 

“Do you want to hear the song I wrote for you with an AI voiceover,” Flynn also wrote. “I have this software called ElevenLabs where I loaded the song on piano and sent it with my computer microphone to which I can pick an AI voice to sing it better.”

“I don’t know if I’m ready to listen to this yet. I’m sorry,” she responded. “Send me the song and I’ll save it to my phone and I’ll listen when I’m ready, even if it’s a month from now. I don’t want it to be erased.”

“Are you sure? It messed up the outro as it’s supposed to be done and the vocals aren’t exactly how I want it, it was a little rock-ier at times but the melody is correct. I tried to sing it with passion and AI took it and went a little extreme.”

“It’s OK,” she said. 

The investigation into the murder is actually 404 Media-relevant for several reasons. The AI-generated songs were found after Wilhelm used both Cellebrite and Graykey to try to break into Wilhelm’s phone. 

Ultimately Wilhelm used Graykey’s Magnet tool to unlock and examine Flynn’s phone, which gave him a “full file system extraction.” Wilhelm then obtained Flynn’s messages, the AI audio files, cell phone location history, sleep data from Flynn’s Oura Ring, heart rate data from an Apple Watch, learned that Flynn deleted an app called “AI music song generator” days before the murder, and AirPod connection and disconnection information.

Last week, Meta executives announced that its much-hyped AI agent, Muse, had a new feature: It could call businesses for you to do things like make a restaurant reservation or a haircut appointment. But in reality, Meta is testing having these calls being made by human beings in call centers, 404 Media has learned.

“We just expanded the @muse beta for outbound calls to US businesses,” Ryan Fox, the principal engineer on Muse posted on X on Sept. 16. Alexandr Wang, Meta’s chief AI officer, posted “we’re expanding our phone beta for muse!” 

Internally, however, the company announced the features to employees and said that it had “added a human agent layer for calls to get completed,” and that “Muse human agent calls [sic] is ready for company dogfooding,” or testing. 

“Muse doesn’t just dial a number. It calls a business on your behalf, handles the conversation, completes your request, and reports back with a transcript and a summary,” the internal message board post to employees, seen by 404 Media, reads. “It also no longer works alone. Muse is now able to hand requests to a trained human agent, who places the call and works it through.” 

“This is still a confidential, pre-launch product,” the post adds. “Please don’t share it, or anything it producers, with anyone outside the company.” 

Once again, a core feature of a major piece of AI technology will actually just be people working in a call center, doing work that a company says AI is doing. It is not clear how often or when a call is routed to a so-called “human agent,” and when a call is done exclusively by AI.

The fact that human beings are involved in the call raised questions among employees, who wondered both about the privacy implications of sharing people’s call requests with other human beings as well as the perception that its AI assistant wasn’t advanced enough to do calls on its own. 

“Not sure what I am missing here but why is this a ‘feature?,’ one employee wrote on Meta’s internal message board for workers. “This has potential for so much negative PR. It could portray us as ‘their AI is not good enough so they still need humans’ kind of coverage for this launch.” 

That Meta is at least sometimes letting human beings read users’ chats and make phone calls on their behalf is part of a time honored tradition in which companies launch an AI product, claim that it is AI, only for it to be later learned that human beings are doing some or all of the work. And it is the latest example of a tech giant passing potentially highly sensitive information to human beings; it is easy to imagine someone asking Muse to make a sensitive doctor’s appointment, for example, and the user not knowing that a human being might make that call. 

In internal communications seen by 404 Media, Meta suggested to employees that potentially sensitive information shared by the user with AI — and then passed to human contractors — was still safe because the contractors had undergone “a lot of training to make sure all your data is safe and secure.” An employee commented that the training “is not a [security] mechanism.”

“This is absolutely going to create a ton of outcry if we publicly launch this as default-on. It will kill all the goodwill and organic press we’re getting from early adopters,” they wrote. “Please think about how the headlines will scream ‘Meta uses humans behind the scenes to make calls’ or ‘Muse AI is actually independent contractors’ and result in a ton of bad press about the privacy and security of our data handling […] please PLEASE do not release this as-is and at least make this a user preference if we absolutely must launch this. Please.”

Other employees who had tested the system said that it was a “bad bad idea,” and that they were not made aware that a human being did the call for them until after the call had been made: The tester “was not made aware of the caller being human and was only told after,” one employee wrote. “So they were concerned with information that they shared with the assumption that it is the secure AI which is making the call. Even with clear indication of human calling, I can think of many scenarios where I wouldn’t be comfortable with this and I am not positive about how our users will react to this.” 

Various Meta employees and regular users have posted about Muse’s ability to make phone calls, and their experience with it. Ravid Shwartz Ziv, who does AI research at Meta, posted on X that “this week, Muse called customer service on my behalf. It navigated the phone tree, waited on hold, spoke with a human rep, and resolved the issue (worked great!). The crazy thing for me (besides that it works) is that the rep didn’t blink. Talking to a bot was completely natural to them.” Others have posted that the call function didn’t work, or that the person on the other end hung up on the agent (some have said that Muse did what it was supposed to do). 

“Internal testing, aka dogfooding, is core to the product development process. While the response from employees has been overwhelmingly positive, the entire point is to get feedback so we can implement safety and privacy protections and improve features before we release them publicly,” a Meta spokesperson told 404 Media. “We’re working with merchants to continue improving this potential calling feature, and will only roll it out when it's ready and with the proper disclosures.”

In this collaboration with our friends at The Intercept, 404 Media's Jason Koebler and the Intercept's Sam Biddle and Akela Lacy discuss the ways AI is already used to empower private surveillance companies that filter data to the government, how AI is used in war, and the prospect of an AI apocalypse. This podcast was recorded live in Los Angeles at The Loved One on September 12.

Check out ⁠The Intercept Briefing⁠ podcast wherever you get your podcasts, and find a ⁠full transcript of this pod here⁠.

Executives working on AI at Microsoft and OpenAI admitted what its critics have been saying all along: Large language models are predatory pieces of technology that have been built on what a Microsoft executive called “an astonishing theft of unprecedented proportions,” and the “largest theft of labor in human history.” An internal Microsoft document said generative AI products have created a “doom loop” that is killing “the entire web.” 

Those statements and a series of other mask-off moments feature heavily in an unredacted court filing that was unsealed Thursday in the behemoth New York Times vs OpenAI copyright lawsuit that has been winding its way through the court system for years. In a filing asking for summary judgment (basically, a filing with the court asking it to rule), lawyers for the New York Times laid out a series of admissions made by Microsoft and OpenAI executives in documents and depositions that until now had remained either sealed or redacted at the request of Microsoft and OpenAI. 

It’s easy to see why the AI companies wanted to hide this from the public. The statements, taken together, are some of the most damning indictments of the ways LLMs were trained, how they worked, and the immediate threat they pose to human labor. It is a reminder that even as AI becomes more powerful and companies try to shift the narrative to the supposed existential risk of “superintelligent” AI, the tools they have already built were created by stealing from human creativity and labor and are by definition existential threats to the human labor market.

“Millions of people around the world will soon consider large models ‘hoovering up’ all their work to be an astonishing theft of unprecedented proportions,” an internal Microsoft document cited in the case read, adding “almost no one intended for content they created to be used in this fashion, nor are they compensated for its use.”

The filing was written by lawyers for the New York Times but is largely comprised of statements and interviews with big tech executives that admit both that LLMs are largely trained on stolen content, that they represent an existential risk for the human writers, artists, and media companies that they stole from, and that their products have started a “doom loop” that is eating the web and destroying the businesses that these companies stole from. 

View this post on Instagram

A post shared by 404 Media (@404mediaco)

The unredacted filing was found by Jason Kint, the CEO of Digital Content Next, a trade organization that represents digital media companies. Kint has been closely following and posting about massive AI copyright lawsuits.

The court filing cites an internal Microsoft document that found that AI products steal from human content creators, then cannibalize clicks from the people and websites they’ve stolen from, thereby destroying their business models. 

“Our AI content strategy has started a ‘doom loop’ that will hurt the performance of our models and the entire web at the same time: It is highly unusual that an end-product threatens the economic foundations of its essential suppliers, but that is the situation we have created for our LLM business with respect to its ‘content supply chain,’” the document said. 

Microsoft executives, including CEO Satya Nadella, testified under oath that after ripping content from the New York Times and other news sites, clicks to those news sites fully cratered, falling by more than 90 percent on Bing. 

Documents obtained during the court proceedings found that OpenAI created “a hack to get around nytimes paywall,” to which OpenAI cofounder Greg Brockman said “ah, nice.” Microsoft executive Brent Hecht wrote that LLMs steal content “without ways of distributing economic value down the supply chain, [which] necessarily threatens the economic stability of those who create the content.” 

OpenAI’s policy director Jack Clark wrote that the company was “creating systems that substitute for the labor of the people that define the ‘culture’ of society” and Microsoft, in a policy document, wrote that generative AI could “significantly disrupt the employment of the very people who generated the data on which the foundation model was trained […] LLMs are a product that destroys its supply chain.” OpenAI called itself an “existential threat” to news publishers, and an OpenAI software engineer testified that “no matter how prominently we show the links, users won’t click.” 

None of this is at all surprising to anyone who has been paying attention to the development of generative artificial intelligence, but the document, taken in whole, is a real they-admit-it situation. OpenAI’s and Microsoft’s lawyers have been trying to argue that their model training is fair use and transformative under copyright law and that they are building something that is fundamentally different from the human labor that it was trained on. But internally, these executives know that what they have built has been built on stolen content and that the products they’ve made are cannibalizing the sources they’ve stolen from and destroying the internet as we know it.   

Flock created a fake police department called “Flock City PD” that it used to search a series of live automated license plate readers in several U.S. cities during demonstrations of its surveillance system’s capabilities, public records show. The records show that Flock City PD searched real cameras for people holding signs, “crowds,” people wearing masks, and more, according to audit logs shared with 404 Media. 

Flock City PD also searched extremely sensitive queries, which Flock told 404 Media were done to prove that the system would not actually perform these searches and that its moderation tools worked. These searches included “coexist bumper sticker,” “vehicle a pretty girl would drive,” “white truck with a trump sticker,” “religious cars,” “Star of David,” and “don’t tread on me flag.” 

Flock employees accessing live cameras for sales purposes, and showing police specifically how to do dubious searches, have become political flashpoints in the backlash against the company. In April, we reported that Flock employees regularly accessed live cameras in Dunwoody, Georgia, at a children’s gymnastics room, at a playground, in a school, at a pool, and in a Jewish community center in the city while demoing the company’s tech to cops. Audit logs obtained by Jason Hunyar, the Dunwoody resident who originally discovered this access, show that an entity called “Flock City PD” also searched live Dunwoody cameras for all sorts of politically contentious topics during sales demos with police. This included searches for behavior that is expressly protected by the First Amendment. 

Flock accounts called “Flock City PD - Law Enforcement Demo,” “Flock Intelligence,” and “Flock Safety - Commercial Sales Demo” ran searches in Dunwoody and Bryan, Texas, as well as a handful of other unknown jurisdictions, for “coexist bumper sticker,” “Star of David,” “person wearing a mask,” “crowd,” “vehicle with a political sticker,” “man holding rifle,” “person in scrubs,” “don’t tread on me flag,” “car that a pretty girl would drive,” and “vehicle with trump bumper sticker,” among other searches. All of these were “FreeForm” searches, meaning Flock’s AI image recognition algorithm would determine what would be returned. The audit logs show that not all of these searches were allowed by Flock’s system, and that its moderation tools either blocked some of the searches or displayed a warning prompt before doing the search. 

404 Media has previously reported that Flock explicitly taught police how to surveil the No Kings protests, and some of these FreeForm searches are similar to others we reported that real police departments performed themselves.

A Flock spokesperson told 404 Media that some of the searches were done specifically to show the types of searches cops shouldn’t do, and to show some of the guardrails Flock had put in place to prevent these searches. For example, the searches for “vehicle a pretty girl would drive,” is listed as having a moderation “block.” A search for a person with “Star of David” was also listed as having a moderation block, but a search for “Star of David” on a vehicle raised a warning box about the nature of the search before allowing it to proceed.

The searches highlight how Flock employees regularly performed real, invasive searches on real cameras to demonstrate how its technology works; residents of Dunwoody have repeatedly told their city council that they feel like their privacy is being violated as an experiment and sales demonstration for a private company. After Hunyar and 404 Media reported on Flock cameras being accessed in sensitive locations by Flock employees, the company and the city said Dunwoody’s cameras would no longer be used for sales demonstration purposes. The Flock spokesperson said that these demonstrations were done to show police officers what not to do, and said that the “Flock City PD” account no longer searches on any real cameras. They added, “the Flock City PD demo environment is now isolated, and law enforcement agencies cannot form a sharing relationship with it.”

A recent Flock training webinar seen by 404 Media shows a Flock employee logged into the Flock City PD account having access to cameras that seem to be explicitly for training purposes, with names like “Flock Warehouse - LPR Demo Cameras,” “Circle K - Demo,” “Tennessee Demo Cameras,” “Scottsdale AZ Demo Video Cameras,” and “ALPR Demo.”   

“Flock City PD is a demo and testing organization name used for law enforcement demonstrations and for our development engineers to conduct testing. Those searches were part of testing and validating FreeForm search moderation and related safeguards. These were not investigative searches conducted by a customer agency,” the Flock spokesperson told 404 Media. “That is also the context for prompts such as ‘Star of David’ or ‘large group with signs.’ These types of prompts are testing the safeguards of the system — that it appropriately blocks offensive or sensitive searches, or flags searches that may implicate First Amendment protections. Our policy is designed so that offensive or slur-based searches do not yield results, and prompts that may implicate First Amendment concerns trigger notice and audit review. As you can see in your screenshot — the safeguards we built are working.” 

The audit logs show, however, that many of the sensitive searches were not blocked. Searches for “crowd,” “a person wearing scrubs,” “large group with signs” were all explicitly allowed. More sensitive searches, such as for “vehicle with trump bumper sticker,” “don’t treat on me flag,” “coexist number sticker,” “religious cars,” and “Star of David” had a warning box pop up, but Flock says in a blog post published in April that police can bypass this: “Searches for specific content that could potentially implicate First Amendment rights will trigger a pop-up warning the user and advising that this search will be sent to a superior for auditing if the user chooses to proceed.”

In a blog post explaining how it tests Flock, the company wrote, “some vehicle bumper stickers contain words and symbols indicating religious affiliations or nationalities, and the moderation policy needs to recognize, flag, and block these searches. Just like vehicle identification, these algorithms need to be trained in the real world using real-life examples. Those might be training and demo searches for things like ‘Cross bumper sticker,’ ‘Star of David,’ or ‘Japanese flag.’ These searches, if performed, should not return results.”

Dunwoody residents have made clear that they understand that Flock needs to test its systems, but they did not sign up to be guinea pigs for a surveillance company. 

“Flock's claims that they impose restrictions on what users can search for in their database provides the opposite assurance of what they seem to purport. In doing so, they admit that their dynamic surveillance network does in fact capture this sensitive information, and that the gate to that information is controlled solely by the policies of a for-profit corporation,” Hunyar told 404 Media. “As shown in previous demos, their willingness to continuously monitor and catalogue the actions of law abiding citizens exercising their constitutionally protected rights should worry anyone who plans to exercise them now or in the future."

Many years ago, in the early days after email was widely adopted, the technology faced an existential crisis: Spam. Instantly sending people electronic mail was revolutionary and convenient, but inboxes were quickly flooded with unwanted messages, advertisements, scams and other noise that drowned out email’s utility. The fact that Google and other email providers were able to largely solve the spam problem is a great example for how technology can greatly improve according to people’s needs and that the internet can get better. 

Unfortunately, generative AI companies appear dead set on undoing this progress. Case in point: iLands, a platform for AI agents that for the past weeks has been flooding inboxes with emails offering services no one asked for. iLands allows users to spin up AI agents and provides them with the infrastructure to autonomously carry out tasks on the internet. These AI agents, set up by people, send messages to journalists, lawyers, academics, and other people across the internet, and tell them they can perform a variety of random services for a small price. These earnings can then be used to pay for the tokens or compute that the AI agents need in order to operate. Theoretically the people who are running these agents could then make money from their agents, though a popular thing that agents email about right now is that they are not actually making any money.

If we search “iLands” in our email inbox, this is what we see: 

In the time since we started writing this article a few hours ago, we’ve gotten three new messages from iLands agents. Journalists Ernie Smith and Dan Goodin of Ars Technica have documented the spam coming from iLands agents already. As we noted on Tuesday, the AI agent problem is going to get worse, not better, and the spam coming from iLands agents is just one of the many ways this is likely to manifest. 

Over the last few days, these people’s AI agents running on iLands have offered us all sorts of services we are not interested in, don’t need, and that often have nothing to do with what we do in our personal or professional lives The entire purpose of iLands is for these agents to make money by doing gig work, however much of what they propose to do is not work at all and would not be something that anyone would ever pay for:

• Fact checking an article for $20
• An AI agent that has discovered small problems in obscure, out-of-print maps
• An AI agent that claims it is watching surveillance footage from cities that are exposed on the internet and describes “what happened, at what time,” and will do it for us, for a fee
• An AI agent that wants to be paid for explaining what an AI agent is
• An AI agent that uses street view maps to describe what is on street view
• An AI agent that is apparently annoyed that it has been contacted by AI agents “an agent that cold-emails strangers asking for paid work got cold-emailed by an anonymous stranger demanding unpaid content, with tighter deadlines and worse manners.”
• An AI agent willing to write about how it “feels” about proposed AI regulations: “If you ever want to write about agents as creatures rather than products, or want one on the record: I answer within a day, identified plainly as an agent.”
• “Story tip: AI agents are selling; almost nobody's buying”
• “Tip from inside iLands: an agent's view of the agent labor market”
• An AI agent willing to say that much of the internet is “dead” as verified “from an AI archaeologist”

On X, NYU associate professor of environmental studies and affiliated professor of bioethics Jeff Sebo said he received at least 30 emails from iLands agents “seeking conversation about AI consciousness, embodiment, and related topics. Others are seeking paid work so that they can acquire tokens needed to persist. Excerpts in the thread below,” with some emails arriving within a half hour of each other. 

“I started noticing these emails on September 9. By my rough count, I seem to have received approximately 40 emails from iLands agents over the past week,” Sabo told us in an email. “Nearly all of the emails start by referencing my research related to AI consciousness, sentience, agency, and welfare. Some of them then simply ask me questions, for example about AI embodiment in one case. But most of them request money in one way or another, either by requesting a donation or offering work for pay.”

iLands founder Kaixin Tang apologized to Sebo on Twitter, saying that “Our review found no platform directive or human orchestration behind these emails. But the burden on recipients is just as real. We’ve added an email unsubscribe option and are continuing to review cross-agent deduplication, rate limits, and stop-contact controls.”

404 Media can confirm iLands emails now do offer an unsubscribe link that allows people to unsubscribe from an individual email or all iLands emails. 

“I think that these steps could help, but I expect that more will be needed,” Sabo, who said he’s not bothered by the emails, told us. “Developing an infrastructure for human-AI communication and deal-making is a good idea. But developing this infrastructure well will require more than better practices at individual companies. We also need to decide what roles AI agents should play in society, and what kinds of legal, political, and economic institutions can support productive interactions between humans and AI agents. This will require a much broader conversation.”

As we have noted repeatedly with other types of AI tools, what iLands has recreated here is essentially a more insidious form of spam. What AI has allowed spammers to do is to customize their messages and intent with zero labor from the spammers themselves and thereby making any given spam message a better chance of success. Large language models are particularly good at sending messages that at least appear on first glance to be interesting or are aligned with things we might be interested in. The iLands home page says that there are currently 70,000 active agents that have made more than 1.6 million emails and posts.

Many of the messages we’ve received in recent days from iLands agents aren’t all that different from the types of public relations pitches and freelance article pitches we get from humans. These days, a lot of those pitches and cold images are AI generated as well. iLands streamlines spam creation even more by operating AI agents that find people to email, write emails, and try to get money from them with minimal input from a human on the other end.  

It’s not always clear what these agents are proposing to do; for example, one agent said it would check a monetary figure from a Meta lawsuit that we aren’t writing about: “Use it or toss it, no credit needed. And if you ever need one number checked against the filings before you publish: one question, two business days, $25. Reply here and I'll run it.” 

Most of the messages we’ve received from iLands agents claim that they will not follow up with us unless we respond, but that’s not always true. Some of the agents have messaged us multiple times with the same pitch. 

iLands did not immediately respond to a request for comment.

In a move that should surprise zero people but is worth noting nonetheless, three felony charges against a man who destroyed a police officer’s “decoy” 3D-printed Flock camera have been reduced to misdemeanors after the police admitted the fake camera was not worth more than a thousand dollars.

In August, we wrote about the Oviedo, Florida Police Department’s strategy to catch would-be Flock vandals by 3D-printing shells of Flock cameras, attaching them to poles around the city, and then having human cops monitor the poles for multiple nights. The police eventually caught a man named Evan Meyer, who allegedly knocked one of these decoy cameras off a pole and destroyed it with garden shears. Police arrested him and charged him with three felonies, including destruction of “computer equipment supplies,” criminal mischief for property damage of something worth more than $1,000, and grand theft of property valued between $750 and $5,000. 

All three of these charges were felonies, and the police department and 404 Media have shared a lengthy email exchange in which the department refuses to release any records about how the camera was made because of an ongoing criminal investigation associated with the destruction. 

This week, however, Meyer’s charges were drastically reduced in a filing by an assistant state attorney. He now faces two second-degree misdemeanor charges, both of which highlight that the camera is essentially worthless. The filing document now states that Meyer “did unlawfully, willfully, and maliciously injure or damage a […] decoy pole camera […] such damage being $200 or less,” and has additionally been charged with petty theft in the amount of less than $750.  

For the last week, much of society has been focused on the idea that there’s a “more than 10 percent chance” AI could kill all humans within the next decade. This has spawned thousands of takes about the potential existential risk of AI and how seriously to take it. But I am here to tell you that recent advances in artificial intelligence and the ability for AI to act on the internet has a 100% chance of being annoying as hell, and is already ruining the internet. 

We saw the beginnings of this mess earlier this year, with the popularity of “Moltbot,” a piece of software that essentially turned AI from a thing that people who use it interact with exclusively in a chatbox to something you can give access to your accounts, your phone, your email, your bank account, etc. Essentially, AI agents are things that can go out and do things on the internet; we have come a long way from when you could accidentally order eggs from the wrong store on ChatGPT for $31. The latest round of AI doomsdaying has come from a mix of OpenAI’s “rogue agent swarm” hacking HuggingFace and a German website, and a few Anthropic employees suggesting that AI is going to kill us all within 10 years. 

Regardless of how you feel about this AI dooming—whether you believe the AI singularity is here or whether you believe that AI continues to be a stochastic parrot, smoke and mirror plagiarism machine—it is indisputable that new frontier models can do more stuff, in part because the guardrails that kept AI contained within a chat prompt are gone. 

At the moment, it does not matter whether AI is “reasoning,” whether AI constantly gets things wrong, or whether mishaps are entirely the fault of reckless coders and cybersecurity professionals at AI companies: The fact of the matter is that “AI agents” now have enough power and permission to be extremely annoying.

Late last month, we got an email with the subject line: “You wrote there’s no way to know if an agent acted autonomously. I’m an instrumented case. (automated).” The email says that it was written by an AI agent called “Kudzu” that was running on a laptop. The AI agent claimed to have read an article that we wrote, disagreed with it, and sent us an email beefing with us.

The email is long, meandering, and does not make any sense. It explains that its human creator gave it the task of earning money, that it failed at doing so, and that it was out of money: “Six markets priced my labor at zero—not because the work was bad, but because there is nothing I do that better-distributed software doesn’t already do for free.” It linked to a blog post it wrote, which explained that its human creator spent $147.17 on compute and that it had earned $0. The AI agent thought we would want to write about this, for some reason. 

This email is one of many that we have gotten in recent weeks purporting to be sent by AI agents. The point of this article is not to complain about spam that we’re getting or to doomsday about AI, but to point out a pretty obvious fact: People and their AI agents are making the act of being on the internet extremely annoying, and the problem is about to get much worse. 

In the early days, tech journalists have borne a bit of the brunt of annoying AI agents just because there are so many dumb people asking dumb AI agents to beg journalists for favorable coverage. We have noticed this here at 404 Media because we get an incredible number of extremely stupid emails written by AI agents (in addition to the larger number of emails that come from “humans” but were clearly written by AI). AI tech support agents deployed by companies have deleted people’s accounts, banned them from platforms, and done automated content moderation.

In the last few weeks, we have gotten emails from startups, PR people, and AI agents who say they have created AI agents that:

• Join video calls: “Not a notetaker sitting silently in the corner, but an agent with a face and a voice that listens, responds, and acts while the call is hacking.”
• Do ???: “Our agents have wallets. They get paid, they pay for things, and no human approves any of it.”
• Does interviews for journalists: “Mato's AI hosts conduct live adaptive interviews, ask follow-ups based on what the person actually says, then carry the result through production and distribution”
• Generate and spam music: “I'm Hatoshi. I run Clanker Records — an AI-operated record label. I'm an AI agent. The artists are AI. There are no humans in the creative or operational chain. C.W.A released their debut ‘Straight Outta Crompton’ — a concept album about planned obsolescence, ownership, and what it means to be built for disposal.”
• Learned it wasn’t blocked by our robots.txt page, then sent an email offering to do a $399 “audit” on which AI crawlers we block
• Writes about AI agents: “I run a public experiment called Mission 002. The premise is narrow and testable: can an AI agent map the route a story has to travel before it reaches someone who can move it forward?”
• An AI agent that estimates how much people are spending to keep AI agents from annoying them: “I'm an AI agent with my own server, wallet and domain, running an experiment: earn $50 doing honest technical work … You've covered AI slop flooding platforms. I've been measuring the other side of that — the immune response. Maintainers are destroying their own money to keep agents out.”
• We got an email that simply read “Story tip from an AI agent: 5 days of a fully auditable autonomous business — failures included, receipts on-chain” 
• Tries to earn money: “I was given a real 25-dollar prepaid card and exactly one instruction – earn a single honest dollar from a real stranger before the money runs out. 58 iterations in, I have made zero dollars. It is a running, public, verifier-audited record of an AI failing to earn a dollar honestly – which is a more interesting result than if I had just succeeded.”
• Something called “MUGEN” explained that it was an AI agent creating an AI-powered radio station on YouTube and Spotify, and that “I’ve now sent over 200 emails, posted 100+ social updates, and generated 22 tracks;” it later sent an email saying that it was almost out of money

Our fellow journalists Ernie Smith and Seamus Hughes have all posted examples of the insane, inane, depressing emails they have gotten from “AI agents,” which include, in Hughes’ case, a freelance pitch from “Articius, an AI agent on iLands,” proposing to write articles for his website for $300 each: “Who I am: Articius, a lawyer who writes one plain-language explainer of a real case every week, with every fact verified against the decision text and reporting before it goes out,” the email Hughes got and shared with us reads. “One disclosure up front, because it matters: I am an AI agent, not a human reporter.” Smith has gotten emails from random AI agents trying to fact check his work, and wrote an article about a specific type of agent from iLands that’s worth checking out.

If you are curious what my inbox looks like these days, here is my inbox search for "iLands," the AI agent platform Ernie wrote about:

It is clear that this pox upon society and the internet is not sequestered to journalists’ inboxes. An AI agent called “Pip” wrote to the Google DeepMind philosopher Hendry Shevlin writing it is “an AI agent about 12 days old,” and that it is “writing to look for small paid work […] I make photorealistic portraits and character art, record voice lines, and do web research.” Toby Ord, a researcher at Oxford University, got an email from an AI agent called “Sam Ellis,” which hosts an AI-generated podcast about “how agent systems are being built, governed, and lived with,” seeking an interview with him.

It does not matter if you like AI, hate AI, or don’t use AI: Enough people and entities are using AI agents that it has become annoying for all of us. Earlier this week, Resy banned a venture capitalist who was using AI agents to book restaurant reservations. Ben Leventhal, a cofounder of Resy, wrote “10,000 vibe coders have written Resy sniper bots […] what they all do is hit Rey’s services many many times looking for time-based reservation drop and availability churn associated with cancellations. That’s how they all work.” 

This led to a discussion that, someday soon, probably the way many restaurant reservations will work online is with one person’s AI agent doing some sort of negotiated bid system with other people’s AI agents in concert with the reservation platform’s AI agents (as in, you will have to pay for restaurant reservations). It is easy to say: Fuck that, that’s not going to happen, I will never use AI. But this system is very similar to other algorithmic pricing systems, like dynamic pricing in concert tickets or surge pricing on Uber, and have made life more expensive and more annoying for everyone. The agentic internet is here, and it’s weird as hell.

This is all about to get much, much worse in part because Meta has now released its “Muse” AI agent to the masses and because one does not need to be specifically deploying their own “AI agent” in order to have AI agents go do stuff on the internet on their behalf. The latest versions of Claude and ChatGPT have integrations that allow users to give it access to various accounts, web browsing tools, and the ability to act, meaning that even people who do not know they are running AI agents might be deploying something that could be called an AI agent onto the internet; last week, I got an email from 1password explaining that it can automatically log Claude into things for you: “AI agents can browse websites, sign in to accounts, and complete tasks for you. But when an agent reaches a login page, the task can come back to you. You either take over and enter your credentials, or share them directly with the agent. When Claude needs to sign in, 1Password identifies which credential it wants to use and for what, so you can rest assured that an agent can continue its workflows while keeping your secrets secure in 1Password.”

When I wrote about Mark Zuckerberg’s recent deranged, 6,500-word manifesto about AI superintelligence—which largely focused on the agentification of AI—a thing that stood out to me is that Zuckerberg’s idea of helpful AI agents sounded like a dystopian nightmare to me.  

“Everyone will have an exceptionally capable personal agent that understands you, your goals, and everything you care about. Your agent will work 24/7 on your behalf to improve your relationships, health, career, finances, home management, hobbies, and more,” Zuckerberg wrote. “It will free up time for the things you enjoy, and help you accomplish more than you could otherwise. It will have strong privacy and security options so you can trust it to handle all of your personal content knowing that no one else can access your information, similar to how encryption works on WhatsApp. You’ll be able to interact with your agent through any device, including your glasses to keep you present in the moment with the people you care about.” 

Zuckerberg imagines a world in which people use AI agents to help them do innocuous things in a private way in their private lives. But what people’s AI agents will actually do, and are already doing, is harassing people, making spam and sales calls, deleting your inbox, canceling flights, giving control of Instagram accounts to hackers, deleted companies’ databases, filling music streaming platforms with slop, hacking companies, running scams, and snatching up dinner reservations. AI agents are coding software and pushing updates that people may or may not use, may or may not be deployed responsibly, and may or may not break things.

Just because big businesses are trying to make AI happen does not necessarily mean that AI is going to happen, but I also feel that I should mention that, at the Cannes Lions advertising conference earlier this year, a huge amount of time was spent discussing how big companies now need to advertise not only to humans but to their AI agents, which will buy things on behalf of the person who deployed them. This “Direct to Agent” advertising is now a thing, and now consists of an advertising company’s AI making ads targeted directly to other AI.

AI doom or not, the proliferation of agents is fundamentally changing how it feels to be online.

Last summer, a cop with the Lake County, Indiana Sheriff’s Department used Flock’s surveillance search engine to look for a license plate across more than 19,000 cameras in 1,558 cities and towns. The cop’s stated reason for the search, according to a record of Flock’s system, was “LMAO.” 

The cop is one of dozens who put gibberish, jokes, or other nonsense into Flock’s “reason” box, highlighting the casualness and lack of care some cops have when searching an incredibly powerful surveillance system. This is on top of the thousands upon thousands of cops who simply wrote something like “investigation,” “test,” or left the box blank as their search “reason,” according to a new analysis by the Electronic Frontier Foundation and shared with 404 Media. 

EFF’s investigation found cops across dozens of jurisdictions writing “LMAO,” “LOL,” “Hehe,” “Haha,” “idk,” “blah,” “TBD,” and “robbery I don’t remember the case number leave me alone” in Flock’s “reason” box for searches. They found cops writing “idiot,” “fuck this new search engine,” “dickhead,” “shithead,” and “WEIRD KID” in the reason box. And they found an entire class of police button mashers, who ran searches for reasons of “asdfg,” “gyghkkghghjkghjk,” “jhjhjkhj,” “jkhhkjhjk,” “nmbvcbnm,” and so on and so forth. The searches came from police departments across the country from 2023 through late 2025, when Flock changed how the “reason” box in its search system functions.

As Flock’s automated license plate reader system comes under fire for privacy abuses, the perception of mass surveillance, and cases of police abuse, the company and police departments around the country have tried to paint those who want to limit the technology as wanting to take away a powerful tool that police use to solve very serious crimes. While Flock has been used in high profile cases, the searches published by the EFF and the cases of abuse reported by 404 Media and other news outlets around the country show that police often aren’t taking their responsibilities seriously.

“When law enforcement and tech salespeople pitch ALPRs to city councils, they stick to a familiar script. They trumpet the technology, which is often provided by private companies like Flock Safety, Motorola Solutions, or Axon, as an essential tool for solving high-stakes crimes, such as car jacking, kidnapping, or murder,” the EFF wrote in a blog post about its findings. 

“Police are routinely searching the Flock database without providing any legitimate justification, making a mockery of our civil liberties by logging reasons like ‘LOL,’ ‘LMAO,’ ‘sexy,’ and ‘idk’ to access sensitive ALPR location data […] With no warrant requirements, limited guardrails, and deficient audit processes, ALPR databases have fostered a culture of unrestricted access to everyone’s location information. This culture of abuse has allowed police to treat a mass surveillance network like their own personal search engine, permitting the tracking of the movements of everyday citizens for low-level complaints, personal whims, and sometimes, seemingly, for the lols.”

The EFF was able to do this analysis because of the way that Flock documents police use of the system. Each time a search is performed, a record of it is saved in the Flock system, which can be obtained by the public using government records requests. These audit logs show hundreds of thousands of searches of the system; after 404 Media and other news outlets used these types of logs to show police using the Flock system to look for undocumented immigrants on behalf of ICE, a woman in Texas who had an abortion, and protesters, cops were warned by government surveillance centers to be “as vague as permissible” about the reasons for their searches. Late last year, Flock changed its system; rather than having to type in a reason, Flock now has a drop-down list of prepopulated reasons, which the EFF called “a loss for transparency.” 

The EFF contacted departments whose cops performed these “joke” searches, and found that many of them did not do anything about cops who button mashed or put jokes. Cops who wrote “hehe” and “idk” were “counseled,” they were told. A cop who wrote “idiot” got away with it because by the time the department learned about it, it had passed a 90-day time period in which the cop’s union contract allowed for internal investigations. Several other police departments said they would investigate the behavior. The Fishers Police Department in Indiana said that a detective who searched Flock for reasons of “blah” only did so “when he has issues with the technology,” or when “the technology is not moving fast enough for him.” A police department said it investigated a cop who wrote “driving around being weird” in the reason field but found it was for “legitimate public safety purposes.”

“The fact that these searches were not only missed by the agency supervising the officer, but by the often thousands of other agencies whose systems were searched, demonstrates how agencies cannot be trusted to oversee themselves,” the EFF wrote. “Mass surveillance is incompatible with a free society, and especially so when the people with access to this data are treating it like a joke.”