404 Media

2025-06-25 13:00:03

Podcast: This Site Unmasks Cops With Facial Recognition

We start this week with Emanuel and Joseph’s coverage of ‘⁠FuckLAPD.com⁠’, a website that uses facial recognition to instantly reveal a LAPD officer’s name and salary. The creator has relaunched their similar tool for identifying ICE employees too. After the break, Jason tells us about a massive AI ruling that opens the way for AI companies to scrape everyone’s art. In the subscribers-only section, our regular contributor Matthew describes all the AI slop in the Iran and Israel conflict, and why it matters.
playlist.megaphone.fm?e=TBIEA9…
Listen to the weekly podcast on Apple Podcasts,Spotify, or YouTube. Become a paid subscriber for access to this episode's bonus content and to power our journalism. If you become a paid subscriber, check your inbox for an email from our podcast host Transistor for a link to the subscribers-only version! You can also add that subscribers feed to your podcast app of choice and never miss an episode that way. The email should also contain the subscribers-only unlisted YouTube link for the extended video version too. It will also be in the show notes in your podcast player.
youtube.com/embed/1eieXQIaALA?…

• ⁠‘⁠⁠FuckLAPD.com⁠⁠’ Lets Anyone Use Facial Recognition to Instantly Identify Cops⁠
• ⁠Judge Rules Training AI on Authors' Books Is Legal But Pirating Them Is Not⁠
• ⁠The AI Slop Fight Between Iran and Israel⁠

The AI Slop Fight Between Iran and Israel

There’s a lack of footage from the conflict and AI-generated content is filling the void.

^{Matthew Gault (404 Media)}

404 Media

2025-06-24 13:43:47

‘FuckLAPD.com’ Lets Anyone Use Facial Recognition to Instantly Identify Cops

A new site, FuckLAPD.com, is using public records and facial recognition technology to allow anyone to identify police officers in Los Angeles they have a picture of. The tool, made by artist Kyle McDonald, is designed to help people identify cops who may otherwise try to conceal their identity, such as covering their badge or serial number.

“We deserve to know who is shooting us in the face even when they have their badge covered up,” McDonald told me when I asked if the site was made in response to police violence during the LA protests against ICE that started earlier this month. “fucklapd.com is a response to the violence of the LAPD during the recent protests against the horrific ICE raids. And more broadly—the failure of the LAPD to accomplish anything useful with over $2B in funding each year.”

“Cops covering up their badges? ID them with their faces instead,” the site, which McDonald said went live this Saturday. The tool allows users to upload an image of a police officer’s face to search over 9,000 LAPD headshots obtained via public record requests. The site says image processing happens on the device, and no photos or data are transmitted or saved on the site. “Blurry, low-resolution photos will not match,” the site says.

“fucklapd.com uses data provided by the City of Los Angeles directly to the public,” McDonald told me in an email. “This data has been provided in response to either public records requests or public records lawsuits. That means all of this information belongs to the public and is a matter of public record. fucklapd.com is not scraping any data.”

In addition to potentially identifying officers by name and serial number, FuckLAPD.com also pulls up a police officer’s salary.

“Surprisingly it [the domain name] only costs $10 a year to exercise my first amendment right to say fucklapd.com,” McDonald said.
playlist.megaphone.fm?p=TBIEA2…
I tested the tools by grabbing an image of a white and bald police officer from an LAPD press conference addressing its use of force during the anti-ICE protests in Los Angeles. I uploaded the image to the site, and within a few seconds the site presented me with nine headshots of officers who could be possible matches, all of them bald white men. The first correctly identified the cop in the image I uploaded.

Clicking “view profile” under the result sent me to the Watch the Watchers site by the Stop LAPD Spying Coalition, a community group based in the Skid Row neighborhood of downtown Los Angeles. “All of the information on this website comes from records that were deliberately made public by the City of Los Angeles in response to either public records requests or public records lawsuits,” the Watch the Watchers site says. “We plan to keep refreshing this data from new public records requests as well as to add other data.” Stop LAPD Spying Coalition is not associated with FuckLAPD.com and did not endorse the site.

McDonald told me that since the site launched, it had around 50,000 visitors, but “Because the analysis happens on-device I have no way of knowing what people are using it for, except for some people who have posted screenshots to Twitter or Instagram,” he said.

In 2018 McDonald made another tool called ICEspy which used hundreds of photos of ICE employees from LinkedIn and does much the same thing as FuckLAPD.com. “This app is designed to highlight and embarrass the organization committing atrocities against refugees and immigrants to the United States,” ICEspy’s website says. That tool originally used a Microsoft API, before Microsoft restricted access to it. McDonald said on X that he recently relaunched the tool to run locally on devices. 404 Media tested ICEspy using images of ICE employees on LinkedIn to verify if the tool worked and each result was incorrect; McDonald indicated on X he was looking for others to re-scrape LinkedIn and update the database.

Over the last few months ICE officers have consistently worn masks, neck gaiters, sunglasses, and baseball caps to shield their identity while often refusing to provide their name or even confirm the agency they belong to. This includes while violently assaulting people, detaining U.S. citizens, and pointing weapons at bystanders, leaving little room for recourse or accountability against the individual agents or the agency.

ICE’s constant use of masks has created a climate where people cannot be sure that the heavily armed group of men coming towards them are really federal agents or not. In Philadelphia, a man pretended to be an ICE agent in order to rob an auto repair shop and zip tie an employee. In Brooklyn, a man posed as an immigration officer before attempting to rape a woman.

ICE claims that assaults against its officers have increased by 413 percent, and use this as the justification for covering their faces. But as Philip Bump showed in the Washington Postthere are still plenty of questions about those numbers and their accuracy. ICE says its officers’ family members have been doxed too.

Neither the LAPD or ICE responded to a request for comment.

Joseph Cox contributed reporting.

Woman says man posed as ICE agent before attempted rape in Brooklyn

A 51-year-old woman was approached by a man who said he was an ICE agent before attempting to sexually assault her.

^{Eyewitness News (ABC7 New York)}

404 Media

2025-06-24 13:00:39

This Queer Online Zine Can Only Be Read Via an Ancient Internet Protocol

Unless you’re living in a ChatGPT hype-bro bubble, it’s a pretty common sentiment these days that the internet is getting shittier. Social media algorithms have broken our brains, AI slop flows freely through Google search results like raw sewage, and tech companies keep telling us that this new status quo is not only inevitable, but Good.

Standing in stark opposition to these trends is New Session, an online literary zine accessed via the ancient-but-still-functional internet protocol Telnet.

Like any other zine, New Session features user-submitted poems, essays, and other text-based art. But the philosophy behind each of its digital pages is anything but orthodox.

“In the face of right-wing politics, climate change, a forever pandemic, and the ever-present hunger of imperialist capitalism, we have all been forced to adapt,” reads the intro to New Session’s third issue, titled Adaptations, which was released earlier this month. “Both you and this issue will change with each viewing. Select a story by pressing the key associated with it in the index. Read it again. Come back to it tomorrow. Is it the same? Are you?”

The digital zine is accessibleon the web via a browser-based Telnet client, or if you’re a purist like me, via the command line. As the intro promises, each text piece changes—adapts—depending on various conditions, like what time of day you access it or how many times you’ve viewed it. Some pieces change every few minutes, while others update every time a user looks at it, like gazing at fish inside a digital aquarium.

How New Session looks on Telnet. Images courtesy Cara Esten Hurtle

Once logged in, the zine’s main menu lists each piece along with the conditions that cause it to change. For example, Natasja Kisstemaker’s “Sanctuary” changes with every viewing, based on the current weather. “Signature,” by Kaia Peacock, updates every time you press a key, slowly revealing more of the piece when you type a letter contained in the text—like a word puzzle on Wheel of Fortune.

Cara Esten Hurtle, an artist and software engineer based in the Bay Area, co-founded New Session in 2021 along with Lo Ferris, while searching for something to do with her collection of retro computers during the early days of the COVID-19 pandemic.

“I realized I’d been carrying around a lot of old computers, and I thought it would be cool to be able to do modern stuff on these things,” Hurtle told 404 Media. “I wanted to make something that was broadly usable across every computer that had ever been made. I wanted to be like, yeah, you can run this on a 1991 Thinkpad someone threw away, or you could run it on your modern laptop.”

If you’re of a certain age, you might remember Telnet as a server-based successor to BBS message boards, the latter of which operated by connecting computers directly. It hearkens back to a slower internet age, where you’d log in maybe once or twice a day to read what’s new. Technically, Telnet predates the internet itself, originally developed as anetworked teletype system in the late ‘60s for the internet’s military precursor, the ARPAnet. Years later, it was officially adopted as one of the earliest internet protocols, and today it remains the oldest application protocol still in use—though mainly by enthusiasts like Hurtle.

New Session intentionally embraces this slower pace, making it more like light-interactive fiction than a computer game. For Hurtle, the project isn’t just retro novelty—it’s a radical rejection of the addictive social media and algorithmic attention-mining that have defined the modern day internet.

New Session viewed on a variety of Hurtle's collection of machines. Photos courtesy Cara Esten Hurtle

“I want it to be something where you don’t necessarily feel like you have to spend a ton of time with it,” said Hurtle. “I want people to come back to it because they’re interested in the stories in the same way you’d come back to a book—not to get your streak on Duolingo.”

I won’t go into too much detail, because discovering how the pieces change is kind of the whole point. But on the whole, reading New Session feels akin to a palette cleanser after a long TikTok binge. Its very design evokes the polar opposite of the hyper-consumerist mindset that brought us infinite scrolls and algorithmic surveillance. The fact that you literally can’t consume it all in one session forces readers to engage with the material more slowly and meaningfully, piquing curiosity and exercising intuition.

At the same time, the zine isn’t meant to be a nostalgic throwback to simpler times. New Session specifically solicits works from queer and trans writers and artists, as a way to reclaim a part of internet history that was creditedalmost entirely to white straight men. But Hurtle says revisiting things like Telnet can also be a way to explore paths not taken, and re-assess ideas that were left in the dustbin of history.

“You have to avoid the temptation to nostalgize, because that’s really dangerous and it just turns you into a conservative boomer,” laughs Hurtle. “But we can imagine what aspects of this we can take and claim for our own. We can use it as a window to understand what’s broken about the current state of the internet. You just can’t retreat to it.”

Projects like New Session make a lot of sense in a time when more people are looking backward to earlier iterations of the internet—not to see where it all went wrong, but to excavate old ideas that could have shaped it in a radically different way, and perhaps still can. It’s a reminder of that hidden, universal truth—to paraphrase the famousDavid Graeber quote—that the internet is a thing we make, and could just as easily make differently.

Untold Stories: Setting the Record Straight on Tech's Racial History

An interview with R. Arvid Nelsen highlights his efforts in uncovering the vast amount of history pertaining to black scientists and experts.

^{Michael Martinez (IEEE Computer Society)}

404 Media

2025-06-23 14:11:40

'A Black Hole of Energy Use': Meta's Massive AI Data Center Is Stressing Out a Louisiana Community

A massive data center for Meta’s AI will likely lead to rate hikes for Louisiana customers, but Meta wants to keep the details under wraps.

Holly Ridge is a rural community bisected by US Highway 80, gridded with farmland, with a big creek—it is literally named Big Creek—running through it. It is home to rice and grain mills and an elementary school and a few houses. Soon, it will also be home to Meta’s massive, 4 million square foot AI data center hosting thousands of perpetually humming servers that require billions of watts of energy to power. And that energy-guzzling infrastructure will be partially paid for by Louisiana residents.

The plan is part of what Meta CEO Mark Zuckerberg said would be “a defining year for AI.” On Threads, Zuckerberg boasted that his company was “building a 2GW+ datacenter that is so large it would cover a significant part of Manhattan,” posting a map of Manhattan along with the data center overlaid. Zuckerberg went on to say that over the coming years, AI “will drive our core products and business, unlock historic innovation, and extend American technology leadership. Let's go build! 💪”

Mark Zuckerberg (@zuck) on Threads
This will be a defining year for AI. In 2025, I expect Meta AI will be the leading assistant serving more than 1 billion people, Llama 4 will become the leading state of the art model, and we’ll build an AI engineer that will start contributing increasing amounts of code to our R&D efforts. To power this, Meta is building a 2GW+ datacenter that is so large it would cover a significant part of Manhattan.
Threads

What Zuckerberg did not mention is that "Let's go build" refers not only to the massive data center but also three new Meta-subsidized, gas power plants and a transmission line to fuel it serviced by Entergy Louisiana, the region’s energy monopoly.

Key details about Meta’s investments with the data center remain vague, and Meta’s contracts with Entergy are largely cloaked from public scrutiny. But what is known is the $10 billion data center has been positioned as an enormous economic boon for the area—one that politicians bent over backward to facilitate—and Meta said it will invest $200 million into “local roads and water infrastructure.”

A January report from NOLA.com said that the the state had rewritten zoning laws, promised to change a law so that it no longer had to put state property up for public bidding, and rewrote what was supposed to be a tax incentive for broadband internet meant to bridge the digital divide so that it was only an incentive for data centers, all with the goal of luring in Meta.

But Entergy Louisiana’s residential customers, who live in one of the poorest regions of the state, will see their utility bills increase to pay for Meta’s energy infrastructure, according to Entergy’s application. Entergy estimates that amount will be small and will only cover a transmission line, but advocates for energy affordability say the costs could balloon depending on whether Meta agrees to finish paying for its three gas plants 15 years from now. The short-term rate increases will be debated in a public hearing before state regulators that has not yet been scheduled.

The Alliance for Affordable Energy called it a “black hole of energy use,” and said “to give perspective on how much electricity the Meta project will use: Meta’s energy needs are roughly 2.3x the power needs of Orleans Parish … it’s like building the power impact of a large city overnight in the middle of nowhere.”

404 Media reached out to Entergy for comment but did not receive a response.

By 2030, Entergy’s electricity prices are projected to increase 90 percent from where they were in 2018, although the company attributes much of that to damage to infrastructure from hurricanes. The state already has a high energy cost burden in part because of a storm damage to infrastructure, and balmy heat made worse by climate change that drives air conditioner use. The state's homes largely are not energy efficient, with many porous older buildings that don’t retain heat in the winter or remain cool in the summer.

“You don't just have high utility bills, you also have high repair costs, you have high insurance premiums, and it all contributes to housing insecurity,” said Andreanecia Morris, a member of Housing Louisiana, which is opposed to Entergy’s gas plant application. She believes Meta’s data center will make it worse. And Louisiana residents have reasons to distrust Entergy when it comes to passing off costs of new infrastructure: in 2018, the company’s New Orleans subsidiary was caught paying actors to testify on behalf of a new gas plant. “The fees for the gas plant have all been borne by the people of New Orleans,” Morris said.

In its application to build new gas plants and in public testimony, Entergy says the cost of Meta’s data center to customers will be minimal and has even suggested Meta’s presence will make their bills go down. But Meta’s commitments are temporary, many of Meta’s assurances are not binding, and crucial details about its deal with Entergy are shielded from public view, a structural issue with state energy regulators across the country.

AI data centers are being approved at a breakneck pace across the country, particularly in poorer regions where they are pitched as economic development projects to boost property tax receipts, bring in jobs and where they’re offered sizable tax breaks. Data centers typically don’t hire many people, though, with most jobs in security and janitorial work, along with temporary construction work. And the costs to the utility’s other customers can remain hidden because of a lack of scrutiny and the limited power of state energy regulators. Many data centers—like the one Meta is building in Holly Ridge—are being powered by fossil fuels. This has led to respiratory illness and other health risks and emitting greenhouse gasses that fuel climate change. In Memphis, a massive data center built to launch a chatbot for Elon Musks’ AI company is powered by smog-spewing methane turbines, in a region that leads the state for asthma rates.

“In terms of how big these new loads are, it's pretty astounding and kind of a new ball game,” said Paul Arbaje, an energy analyst with the Union of Concerned Scientists, which is opposing Entergy’s proposal to build three new gas-powered plants in Louisiana to power Meta’s data center.

Entergy Louisiana submitted a request to the state’s regulatory body to approve the construction of the new gas-powered plants that would create 2.3 gigawatts of power and cost $3.2 billion in the 1440 acre Franklin Farms megasite in Holly Ridge, an unincorporated community of Richland Parish. It is the first big data center announced since Louisiana passed large tax breaks for data centers last summer.

In its application to the public utility commission for gas plants, Entergy says that Meta has a planned investment of $5 billion in the region to build the gas plants in Richland Parish, Louisiana, where it claims in its application that the data center will employ 300-500 people with an average salary of $82,000 in what it points out is “a region of the state that has long struggled with a lack of economic development and high levels of poverty.” Meta’s official projection is that it will employ more than 500 people once the data center is operational. Entergy plans for the gas plants to be online by December 2028.

In testimony, Entergy officials refused to answer specific questions about job numbers, saying that the numbers are projections based on public statements from Meta.

A spokesperson for Louisiana’s Economic Development told 404 Media in an email that Meta “is contractually obligated to employ at least 500 full-time employees in order to receive incentive benefits.”

When asked about jobs, Meta pointed to a public facing list of its data centers, many of which the company says employ more than 300 people. A spokesperson said that the projections for the Richland Parish site are based on the scale of the 4 million square foot data center. The spokesperson said the jobs will include “engineering and other technical positions to operational roles and our onsite culinary staff.”

When asked if its job commitments are binding, the spokesperson declined to answer, saying, “We worked closely with Richland Parish and Louisiana Economic Development on mutually beneficial agreements that will support long-term growth in the area.”

Others are not as convinced. “Show me a data center that has that level of employment,” says Logan Burke, executive director of the Alliance for Affordable Energy in Louisiana.

Entergy has argued the new power plants are necessary to satiate the energy need from Meta’s massive hyperscale data center, which will be Meta’s largest data center and potentially the largest data center in the United States. It amounts to a 25 percent increase in Entergy Louisiana’s current load, according to the Alliance for Affordable Energy.

Entergy requested an exemption from a state law meant to ensure that it develops energy at the lowest cost by issuing a public request for proposals, claiming in its application and testimony that this would slow them down and cause them to lose their contracts with Meta.

Meta has agreed to subsidize the first 15 years of payments for construction of the gas plants, but the plant’s construction is being financed over 30 years. At the 15 year mark, its contract with Entergy ends. At that point, Meta may decide it doesn’t need three gas plants worth of energy because computing power has become more efficient or because its AI products are not profitable enough. Louisiana residents would be stuck with the remaining bill.

“It's not that they're paying the cost, they're just paying the mortgage for the time that they're under contract,” explained Devi Glick, an electric utility analyst with Synapse Energy.

When asked about the costs for the gas plants, a Meta spokesperson said, “Meta works with our utility partners to ensure we pay for the full costs of the energy service to our data centers.” The spokesperson said that any rate increases will be reviewed by the Louisiana Public Service Commission. These applications, called rate cases, are typically submitted by energy companies based on a broad projection of new infrastructure projects and energy needs.

Meta has technically not finalized its agreement with Entergy but Glick believes the company has already invested enough in the endeavor that it is unlikely to pull out now. Other companies have been reconsidering their gamble on AI data centers: Microsoft reversed course on centers requiring a combined 2 gigawatts of energy in the U.S. and Europe. Meta swept in to take on some of the leases, according to Bloomberg.

And in the short-term, Entergy is asking residential customers to help pay for a new transmission line for the gas plants at a cost of more than $500 million, according to Entergy’s application to Louisiana’s public utility board. In its application, the energy giant said customers’ bills will only rise by $1.66 a month to offset the costs of the transmission lines. Meta, for its part, said it will pay up to $1 million a year into a fund for low-income customers. When asked about the costs of the new transmission line, a Meta spokesperson said, “Like all other new customers joining the transmission system, one of the required transmission upgrades will provide significant benefits to the broader transmission system. This transmission upgrade is further in distance from the data center, so it was not wholly assigned to Meta.”

When Entergy was questioned in public testimony on whether the new transmission line would need to be built even without Meta’s massive data center, the company declined to answer, saying the question was hypothetical.

Some details of Meta’s contract with Entergy have been made available to groups legally intervening in Entergy’s application, meaning that they can submit testimony or request data from the company. These parties include the Alliance for Affordable Energy, the Sierra Club and the Union of Concerned Scientists.

But Meta—which will become Entergy’s largest customer by far and whose presence will impact the entire energy grid—is not required to answer questions or divulge any information to the energy board or any other parties. The Alliance for Affordable Energy and Union of Concerned Scientists attempted to make Meta a party to Entergy’s application—which would have required it to share information and submit to questioning—but a judge denied that motion on April 4.

The public utility commissions that approve energy infrastructure in most states are the main democratic lever to assure that data centers don’t negatively impact consumers. But they have no oversight over the tech companies running the data centers or the private companies that build the centers, leaving residential customers, consumer advocates and environmentalists in the dark. This is because they approve the power plants that fuel the data centers but do not have jurisdiction over the data centers themselves.

“This is kind of a relic of the past where there might be some energy service agreement between some large customer and the utility company, but it wouldn't require a whole new energy facility,” Arbaje said.

A research paper by Ari Peskoe and Eliza Martin published in March looked at 50 regulatory cases involving data centers, and found that tech companies were pushing some of the costs onto utility customers through secret contracts with the utilities. The paper found that utilities were often parroting rhetoric from AI boosting politicians—including President Biden—to suggest that pushing through permitting for AI data center infrastructure is a matter of national importance.

“The implication is that there’s no time to act differently,” the authors wrote.

In written testimony sent to the public service commission, Entergy CEO Phillip May argued that the company had to bypass a legally required request for proposals and requirement to find the cheapest energy sources for the sake of winning over Meta.

“If a prospective customer is choosing between two locations, and if that customer believes that location A can more quickly bring the facility online than location B, that customer is more likely to choose to build at location A,” he wrote.

Entergy also argues that building new gas plants will in fact lower electricity bills because Meta, as the largest customer for the gas plants, will pay a disproportionate share of energy costs. Naturally, some are skeptical that Entergy would overcharge what will be by far their largest customer to subsidize their residential customers. “They haven't shown any numbers to show how that's possible,” Burke says of this claim. Meta didn’t have a response to this specific claim when asked by 404 Media.

Some details, like how much energy Meta will really need, the details of its hiring in the area and its commitment to renewables are still cloaked in mystery.

“We can't ask discovery. We can't depose. There's no way for us to understand the agreement between them without [Meta] being at the table,” Burke said.

It’s not just Entergy. Big energy companies in other states are also pushing out costly fossil fuel infrastructure to court data centers and pushing costs onto captive residents. In Kentucky, the energy company that serves the Louisville area is proposing 2 new gas plants for hypothetical data centers that have yet to be contracted by any tech company. The company, PPL Electric Utilities, is also planning to offload the cost of new energy supply onto its residential customers just to become more competitive for data centers.

“It's one thing if rates go up so that customers can get increased reliability or better service, but customers shouldn't be on the hook to pay for new power plants to power data centers,” Cara Cooper, a coordinator with Kentuckians for Energy Democracy, which has intervened on an application for new gas plants there.

These rate increases don’t take into account the downstream effects on energy; as the supply of materials and fuel are inevitably usurped by large data center load, the cost of energy goes up to compensate, with everyday customers footing the bill, according to Glick with Synapse.

Glick says Entergy’s gas plants may not even be enough to satisfy the energy needs of Meta’s massive data center. In written testimony, Glick said that Entergy will have to either contract with a third party for more energy or build even more plants down the line to fuel Meta’s massive data center.

To fill the gap, Entergy has not ruled out lengthening the life of some of its coal plants, which it had planned to close in the next few years. The company already pushed back the deactivation date of one of its coal plants from 2028 to 2030.

The increased demand for gas power for data centers has already created a widely-reported bottleneck for gas turbines, the majority of which are built by 3 companies. One of those companies, Siemens Energy, told Politico that turbines are “selling faster than they can increase manufacturing capacity,” which the company attributed to data centers.

Most of the organizations concerned about the situation in Louisiana view Meta’s massive data center as inevitable and are trying to soften its impact by getting Entergy to utilize more renewables and make more concrete economic development promises.

Andreanecia Morris, with Housing Louisiana, believes the lack of transparency from public utility commissions is a bigger problem than just Meta. “Simply making Meta go away, isn't the point,” Morris says. “The point has to be that the Public Service Commission is held accountable.”

Burke says Entergy owns less than 200 megawatts of renewable energy in Louisiana, a fraction of the fossil fuels it is proposing to fuel Meta’s center. Entergy was approved by Louisiana’s public utility commission to build out three gigawatts of solar energy last year , but has yet to build any of it.

“They're saying one thing, but they're really putting all of their energy into the other,” Burke says.

New gas plants are hugely troubling for the climate. But ironically, advocates for affordable energy are equally concerned that the plants will lie around disused - with Louisiana residents stuck with the financing for their construction and upkeep. Generative AI has yet to prove its profitability and the computing heavy strategy of American tech companies may prove unnecessary given less resource intensive alternatives coming out of China.

“There's such a real threat in such a nascent industry that what is being built is not what is going to be needed in the long run,” said Burke. “The challenge remains that residential rate payers in the long run are being asked to finance the risk, and obviously that benefits the utilities, and it really benefits some of the most wealthy companies in the world, But it sure is risky for the folks who are living right next door.”

The Alliance for Affordable Energy expects the commission to make a decision on the plants this fall.

Want to build a gas plant? Get in line. - E&E News by POLITICO

A program in Texas shows how political enthusiasm can’t escape the limits of the supply chain.

^{Jason Plautz (E&E News by POLITICO)}

404 Media

2025-06-18 13:00:15

Podcast: Airlines Sold Your Flight Data to DHS—And Covered It Up

This week we start with Joseph’s article about the U.S’s major airlines selling customers’ flight information to Customs and Border Protection and then telling the agency to not reveal where the data came from. After the break, Emanuel tells us how AI scraping bots are breaking open libraries, archives, and museums. In the subscribers-only section, Jason explains the casual surveillance relationship between ICE and local cops, according to emails he got.
playlist.megaphone.fm?e=TBIEA4…
Listen to the weekly podcast on Apple Podcasts,Spotify, or YouTube. Become a paid subscriber for access to this episode's bonus content and to power our journalism. If you become a paid subscriber, check your inbox for an email from our podcast host Transistor for a link to the subscribers-only version! You can also add that subscribers feed to your podcast app of choice and never miss an episode that way. The email should also contain the subscribers-only unlisted YouTube link for the extended video version too. It will also be in the show notes in your podcast player.
youtube.com/embed/Auc7NPD2ig4?…

• Airlines Don't Want You to Know They Sold Your Flight Data to DHS
• AI Scraping Bots Are Breaking Open Libraries, Archives, and Museums
• Emails Reveal the Casual Surveillance Alliance Between ICE and Local Police

Emails Reveal the Casual Surveillance Alliance Between ICE and Local Police

Police departments in Oregon created an "analyst group" where they casually offer each other assistance with surveillance tools.

^{Jason Koebler (404 Media)}

404 Media

2025-06-17 14:47:09

Meta Invents New Way to Humiliate Users With Feed of People's Chats With AI

I was sick last week, so I did not have time to write about the Discover Tab in Meta’s AI app, which, as Katie Notopoulos of Business Insider has pointed out, is the “saddest place on the internet.” Many very good articles have already been written about it, and yet, I cannot allow its existence to go unremarked upon in the pages of 404 Media.

If you somehow missed this while millions of people were protesting in the streets, state politicians were being assassinated, war was breaking out between Israel and Iran, the military was deployed to the streets of Los Angeles, and a Coinbase-sponsored military parade rolled past dozens of passersby in Washington, D.C., here is what the “Discover” tab is: The Meta AI app, which is the company’s competitor to the ChatGPT app, is posting users’ conversations on a public “Discover” page where anyone can see the things that users are asking Meta’s chatbot to make for them.

This includes various innocuous image and video generations that have become completely inescapable on all of Meta’s platforms (things like “egg with one eye made of black and gold,” “adorable Maltese dog becomes a heroic lifeguard,” “one second for God to step into your mind”), but it also includes entire chatbot conversations where users are seemingly unknowingly leaking a mix of embarrassing, personal, and sensitive details about their lives onto a public platform owned by Mark Zuckerberg. In almost all cases, I was able to trivially tie these chats to actual, real people because the app uses your Instagram or Facebook account as your login.

In several minutes last week, I saved a series of these chats into a Slack channel I created and called “insanemetaAI.” These included:

• entire conversations about “my current medical condition,” which I could tie back to a real human being with one click
• details about someone’s life insurance plan
• “At a point in time with cerebral palsy, do you start to lose the use of your legs cause that’s what it’s feeling like so that’s what I’m worried about”
• details about a situationship gone wrong after a woman did not like a gift
• an older disabled man wondering whether he could find and “afford” a young wife in Medellin, Colombia on his salary (“I'm at the stage in my life where I want to find a young woman to care for me and cook for me. I just want to relax. I'm disabled and need a wheelchair, I am severely overweight and suffer from fibromyalgia and asthma. I'm 5'9 280lb but I think a good young woman who keeps me company could help me lose the weight.”)
• “What counties [sic] do younger women like older white men? I need details. I am 66 and single. I’m from Iowa and am open to moving to a new country if I can find a younger woman.”
• “My boyfriend tells me to not be so sensitive, does that affect him being a feminist?”

Rachel Tobac, CEO of Social Proof Security, compiled a series of chats she saw on the platform and messaged them to me. These are even crazier and include people asking “What cream or ointment can be used to soothe a bad scarring reaction on scrotum sack caused by shaving razor,” “create a letter pleading judge bowser to not sentence me to death over the murder of two people” (possibly a joke?), someone asking if their sister, a vice president at a company that “has not paid its corporate taxes in 12 years,” could be liable for that, audio of a person talking about how they are homeless, and someone asking for help with their cancer diagnosis, someone discussing being newly sexually interested in trans people, etc.

Tobac gave me a list of the types of things she’s seen people posting in the Discover feed, including people’s exact medical issues, discussions of crimes they had committed, their home addresses, talking to the bot about extramarital affairs, etc.

“When a tool doesn’t work the way a person expects, there can be massive personal security consequences,” Tobac told me.

“Meta AI should pause the public Discover feed,” she added. “Their users clearly don’t understand that their AI chat bot prompts about their murder, cancer diagnosis, personal health issues, etc have been made public. [Meta should have] ensured all AI chat bot prompts are private by default, with no option to accidentally share to a social media feed. Don’t wait for users to accidentally post their secrets publicly. Notice that humans interact with AI chatbots with an expectation of privacy, and meet them where they are at. Alert users who have posted their prompts publicly and that their prompts have been removed for them from the feed to protect their privacy.”

Since several journalists wrote about this issue, Meta has made it clearer to users when interactions with its bot will be shared to the Discover tab. Notopoulos reported Monday that Meta seemed to no longer be sharing text chats to the Discover tab. When I looked for prompts Monday afternoon, the vast majority were for images. But the text prompts were back Tuesday morning, including a full audio conversation of a woman asking the bot what the statute of limitations are for a woman to press charges for domestic abuse in the state of Indiana, which had taken place two minutes before it was shown to me. I was also shown six straight text prompts of people asking questions about the movie franchise John Wick, a chat about “exploring historical inconsistencies surrounding the Holocaust,” and someone asking for advice on “anesthesia for obstetric procedures.”

I was also, Tuesday morning, fed a lengthy chat where an identifiable person explained that they are depressed: “just life hitting me all the wrong ways daily.” The person then left a comment on the post “Was this posted somewhere because I would be horrified? Yikes?”

Several of the chats I saw and mentioned in this article are now private, but most of them are not. I can imagine few things on the internet that would be more invasive than this, but only if I try hard. This is like Google publishing your search history publicly, or randomly taking some of the emails you send and publishing them in a feed to help inspire other people on what types of emails they too could send. It is like Pornhub turning your searches or watch history into a public feed that could be trivially tied to your actual identity. Mistake or not, feature or not (and it’s not clear what this actually is), it is crazy that Meta did this; I still cannot actually believe it.

In an industry full of grifters and companies hell-bent on making the internet worse, it is hard to think of a more impactful, worse actor than Meta, whose platforms have been fully overrun with viral AI slop, AI-powered disinformation, AI scams, AI nudify apps, and AI influencers and whose impact is outsized because billions of people still use its products as their main entry point to the internet. Meta has shown essentially zero interest in moderating AI slop and spam and as we have reported many times, literally funds it, sees it as critical to its business model, and believes that in the future we will all have AI friends on its platforms. While reporting on the company, it has been hard to imagine what rock bottom will be, because Meta keeps innovating bizarre and previously unimaginable ways to destroy confidence in social media, invade people’s privacy, and generally fuck up its platforms and the internet more broadly.

If I twist myself into a pretzel, I can rationalize why Meta launched this feature, and what its idea for doing so is. Presented with an empty text box that says “Ask Meta AI,” people do not know what to do with it, what to type, or what to do with AI more broadly, and so Meta is attempting to model that behavior for people and is willing to sell out its users’ private thoughts to do so. I did not have “Meta will leak people’s sad little chats with robots to the entire internet” on my 2025 bingo card, but clearly I should have.

‘AI Instagram Influencers’ Are Deepfaking Their Faces Onto Real Women’s Bodies

AI 'influencers' on Instagram have racked up hundreds of thousands of followers and millions of views by stealing reels from real women to make the AI seem more "believable."

^{Jason Koebler (404 Media)}

404 Media

2025-06-11 16:45:52

Wikipedia Pauses AI-Generated Summaries After Editor Backlash

The Wikimedia Foundation, the nonprofit organization which hosts and develops Wikipedia, has paused an experiment that showed users AI-generated summaries at the top of articles after an overwhelmingly negative reaction from the Wikipedia editors community.

“Just because Google has rolled out its AI summaries doesn't mean we need to one-up them, I sincerely beg you not to test this, on mobile or anywhere else,” one editor said in response to Wikimedia Foundation’s announcement that it will launch a two-week trial of the summaries on the mobile version of Wikipedia. “This would do immediate and irreversible harm to our readers and to our reputation as a decently trustworthy and serious source. Wikipedia has in some ways become a byword for sober boringness, which is excellent. Let's not insult our readers' intelligence and join the stampede to roll out flashy AI summaries. Which is what these are, although here the word ‘machine-generated’ is used instead.”

Two other editors simply commented, “Yuck.”

For years, Wikipedia has been one of the most valuable repositories of information in the world, and a laudable model for community-based, democratic internet platform governance. Its importance has only grown in the last couple of years during the generative AI boom as it’s one of the only internet platforms that has not been significantly degraded by the flood of AI-generated slop and misinformation. As opposed to Google, which since embracing generative AI has instructed its users to eat glue, Wikipedia’s community has kept its articles relatively high quality. As I recently reported last year, editors are actively working to filter out bad, AI-generated content from Wikipedia.

A page detailing the the AI-generated summaries project, called “Simple Article Summaries,” explains that it was proposed after a discussion at Wikimedia’s 2024 conference, Wikimania, where “Wikimedians discussed ways that AI/machine-generated remixing of the already created content can be used to make Wikipedia more accessible and easier to learn from.” Editors who participated in the discussion thought that these summaries could improve the learning experience on Wikipedia, where some article summaries can be quite dense and filled with technical jargon, but that AI features needed to be cleared labeled as such and that users needed an easy to way to flag issues with “machine-generated/remixed content once it was published or generated automatically.”

In one experiment where summaries were enabled for users who have the Wikipedia browser extension installed, the generated summary showed up at the top of the article, which users had to click to expand and read. That summary was also flagged with a yellow “unverified” label.
An example of what the AI-generated summary looked like.
Wikimedia announced that it was going to run the generated summaries experiment on June 2, and was immediately met with dozens of replies from editors who said “very bad idea,” “strongest possible oppose,” Absolutely not,” etc.

“Yes, human editors can introduce reliability and NPOV [neutral point-of-view] issues. But as a collective mass, it evens out into a beautiful corpus,” one editor said. “With Simple Article Summaries, you propose giving one singular editor with known reliability and NPOV issues a platform at the very top of any given article, whilst giving zero editorial control to others. It reinforces the idea that Wikipedia cannot be relied on, destroying a decade of policy work. It reinforces the belief that unsourced, charged content can be added, because this platforms it. I don't think I would feel comfortable contributing to an encyclopedia like this. No other community has mastered collaboration to such a wondrous extent, and this would throw that away.”

A day later, Wikimedia announced that it would pause the launch of the experiment, but indicated that it’s still interested in AI-generated summaries.

“The Wikimedia Foundation has been exploring ways to make Wikipedia and other Wikimedia projects more accessible to readers globally,” a Wikimedia Foundation spokesperson told me in an email. “This two-week, opt-in experiment was focused on making complex Wikipedia articles more accessible to people with different reading levels. For the purposes of this experiment, the summaries were generated by an open-weight Aya model by Cohere. It was meant to gauge interest in a feature like this, and to help us think about the right kind of community moderation systems to ensure humans remain central to deciding what information is shown on Wikipedia.”

“It is common to receive a variety of feedback from volunteers, and we incorporate it in our decisions, and sometimes change course,” the Wikimedia Foundation spokesperson added. “We welcome such thoughtful feedback — this is what continues to make Wikipedia a truly collaborative platform of human knowledge.”

“Reading through the comments, it’s clear we could have done a better job introducing this idea and opening up the conversation here on VPT back in March,” a Wikimedia Foundation project manager said. VPT, or “village pump technical,” is where The Wikimedia Foundation and the community discuss technical aspects of the platform. “As internet usage changes over time, we are trying to discover new ways to help new generations learn from Wikipedia to sustain our movement into the future. In consequence, we need to figure out how we can experiment in safe ways that are appropriate for readers and the Wikimedia community. Looking back, we realize the next step with this message should have been to provide more of that context for you all and to make the space for folks to engage further.”

The project manager also said that “Bringing generative AI into the Wikipedia reading experience is a serious set of decisions, with important implications, and we intend to treat it as such, and that “We do not have any plans for bringing a summary feature to the wikis without editor involvement. An editor moderation workflow is required under any circumstances, both for this idea, as well as any future idea around AI summarized or adapted content.”

The Editors Protecting Wikipedia from AI Hoaxes

WikiProject AI Cleanup is protecting Wikipedia from the same kind of misleading AI-generated information that has plagued the rest of the internet.

^{Emanuel Maiberg (404 Media)}