Disenchanted and disturbed by Google’s military contracts with the U.S. and Israel, Liguori says she could no longer stand working at a company that was going to help kill people or help build technology that “automatically makes it easier to escalate the brutality of war.”

Related

The Pentagon Asked OpenAI for Artificial Intelligence Designed to Rarely Say No

Liguori joined DeepMind in 2023 and began working on AI systems aimed at more efficiently routing power through the electrical grid, a peacetime engineering problem. But as she learned of Google’s Project Nimbus contract, which furnishes Israel’s military and top weapons manufacturers with a suite of advanced artificial intelligence tools, she began to question her role in the company, and the company’s role in the world. The day after Google announced its classified military AI contract with the Pentagon in April, Liguori staged a personal sick-out, using a combination of sick leave and holiday time off to cease her work.

In a letter she sent internally informing DeepMind and her colleagues of her resignation, Liguori urged them to “stay human” in the face of military work. “I no longer believe that our technology will have a net positive impact on society. I think there’s a fantasy world where AI can help decarbonize grids, cut waste, protect ecosystems, and benefit humanity,” she wrote. “But it’s not this world, and it’s not this type of general purpose AI we are racing to develop. Racing to deploy agentic AI in this society and historical moment is a terrible idea, and will definitely be a net evil.”

The Intercept spoke to Liguori, an Italian national living in London, where DeepMind is headquartered, about how she came to view her employer as dangerously powerful. During a series of interviews by phone and via text, which have been edited and condensed for length and clarity, she explained her moral opposition to military work, the role of unionization in dissent, and what hope engineers and scientists might have for influencing their companies.

What was it about DeepMind that made you want to work there?

I had a high opinion of DeepMind; it was a dream company. It had this aura of being a research-oriented place where you could sit close not just to computer scientists, but real scientists and researchers and people with backgrounds in biology, in physics, in mathematics, in ecology, in anthropology.

You joined in 2023, by which point the company had already been involved in Project Nimbus. Were you aware of that project at the time?

Not really. The first year when I joined, I was very new and very isolated in my team. Google DeepMind is a huge company and if you’re not in the right channels, you end up being very siloed. I didn’t know about Project Nimbus, but when the Gaza war escalated, I discovered it. I immediately looked for some internal resistance movement to join, but I didn’t find anything.

I discovered [Google employee advocacy group] No Tech for Apartheid, but I thought it was mainly a U.S. organization. I didn’t know how to connect with colleagues that share concerns, so the only thing that I did was buy a Palestine hoodie and go to the office with that.

Related

Israeli Weapons Firms Required to Buy Cloud Services From Google and Amazon

In the summer of 2024, I went on maternity leave because my daughter was born, and during maternity leave I was watching news from Gaza every day. I was really suffering experiencing maternity and watching a livestreamed genocide, and every time I was thinking of my company’s complicity in that. I thought about not coming back to work at all.

When I was due to come back, I met a colleague that had basically taken my role in my team before maternity leave, and they were actually leaving for ethical reasons. They put me in contact with the union movement and No Tech for Apartheid. And so I found other colleagues fighting against this.

How and when did you learn of the company’s contracts with the U.S. military?

By the time they started engaging with the U.S. Department of War, [Editor’s note: July 2025] I was already involved in activism. By this winter we saw the GenAI.mil contract materializing, and for me, this was a very clear indication of the company’s intentions. At the same time Google ramped up the security level authorization for Cloud: IL6, which is the level required to handle classified workloads.

Related

AI Almost Started a U.S.–China War — and No One Seems to Care

The combination of the two — signing this non-classified deal with the Department of War, and ramping up security level — really meant that we wanted to use AI for war, on the battlefield. Even if the internal narrative was, “No, don’t worry, it’s just for administrative purposes, just to process documents.”

What was happening internally?

Internally there was a lot of conversation in specific channels where people were complaining, but leadership never really answered anything. I had some of my posts deleted.

What had you written, and did you ever get any sort of answer as to why this stuff was deleted?

I had several posts deleted because they contained the word “genocide,” and I was told that genocide is an intrinsically violent word, and so I was violating internal policies. Even when my posts weren’t deleted, I often had conversations with my manager and HR telling me that my tone was not appropriate because I was too aggressive and that I should raise my concerns in different ways.

They basically wanted us to send [internal HR] tickets to raise our concerns. The problem is that when you do that, you receive semi-automated answers with absolutely no content, nobody is engaged on the concern you have raised. There’s no discussion at all, you don’t even interact with a human. It’s a lot of administrative and bureaucratic steps you have to follow to raise concerns. What I wanted was discussion and what I wanted was very clear answers from leadership.

They tell you this is not the place to raise concerns, try to go through this other channel. This is the common technique. They always tell you to do something else, try to make you tired, and you will never get an answer.

[Editor’s note: The Intercept asked Google about Liguori’s criticism of both the contract and how discussion is handled internally. Spokesperson Tom Hewett responded: “We work with public sector organisations worldwide and follow a consistent process for reviewing and entering these contracts, including compliance with our acceptable use policies and AI principles. We also have dedicated channels for employees to safely raise any concerns – ranging from business ethics to misconduct – and have strict workplace policies in place to prevent retaliation of any kind.”]

I think the reason why you would be opposed to Project Nimbus is pretty straightforward: genocide. What were your specific concerns about or opposition to the Department of War contract?

I’m against the use of AI for targeting and in warfare in general, not only for autonomous weapons. I’m against AI for war, and I don’t think a commercial company like Google should have anything to do with that.

AI for targeting is very dangerous; it creates a system where it’s easier to escalate the brutality of war. The first problem is the opacity introduced in the definition of targets and the dehumanization that follows. In principle, if you have to define targets, if you want to define Hamas combatants, you have to think what it actually means and come up with a definition.

But with these systems, you can bypass the definition. The system itself comes out with a definition. You don’t even question it, you just blindly rely on the definition as the ultimate truth. The person just becomes a set of aggregated data points. All individuality is lost, all humanity is lost in in this process.

Related

Facing Years in Prison for Drone Leak, Daniel Hale Makes His Case Against U.S. Assassination Program

It becomes a system that dehumanizes people. It becomes a rubber stamp because the human brain cannot process all this information. You keep a person that formally approves it, but they don’t, they cannot, meaningfully evaluate the output. This links to the accountability gap: Everyone in the kill chain can plausibly say that they didn’t decide.

What did you do to protest the Department of War contract internally?

Mainly messaging on corporate channels to inform colleagues, and speaking with them in person. I tried to channel the widespread dissatisfaction and indignation over the contract into the collective movement we are building, into our union effort. [Editor’s note: DeepMind employees voted to unionize in May.] Many people still don’t know there are other colleagues organizing against these contracts and for a more ethical Google DeepMind.

Related

The AI Doomsday Future Is Not Inevitable

When you are alone and you lose faith in your company, and you stop believing your work makes a positive difference, you can only become disillusioned and frustrated, and you ultimately end in acceptance and cynicism, or you leave. Finding you are not alone, and joining a collective movement, gives a much better option. It gave me the strength to stay, at least for a year.

I also co-organized a demonstration in February in front of the office. The scope was broader than just the Department of War, but the December GenAI.mil launch had given me a deep sense of urgency, and that’s why I organized the demonstration. I knew something like the classified deal was coming, and I knew this meant legitimizing Nimbus-style collaborations to become the standard. The demonstration was for Gaza and for any use of AI in war. We had banners with “No AI in the kill chain” and “No Gemini for Genocide.”

Has your work as a research engineer contributed to your concerns about this technology being used to kill? How has your technical knowledge informed your opposition?

Understanding how these systems work makes it easier to see where the problem is. Understanding statistical aggregation, and how easy it is to just delegate responsibility, delegate decision-making, makes a difference. But I think it’s more important to do some philosophical reasoning here.

Do you have concerns about the reliability of the technology?

I am concerned about unreliability. But I don’t think you should ever look at unreliability separate from a discussion of why AI in warfare is wrong by itself.

A lot of the time what I hear is a focus on error, for example, the Minab school attack. If we focus only on the error, we are basically saying that if there was no error, these systems for automated targeting are OK.

Were there people within Deep Mind who were supportive of this contract or military work generally?

Yes, it was actually a very debated topic. I think there is broad consensus on the Anthropic red lines, which are fully autonomous weapons and mass surveillance.

Related

Pete Hegseth’s Military of the Future: Defined by AI and an Utter Lack of Oversight

But on whether to use AI for military purposes, this is a much more debated topic, and a lot of people buy the narrative of us-versus-them: That we need to defend ourselves against Russia, against China, and so on.

What was your thought process on deciding to leave the company over this?

It was a very challenging year. I came back as a new mom with a full-time job, a baby to handle, I had some health issues, and I was committing to activism basically on the side. Many people told me, you can slow down, but I couldn’t. I couldn’t stay in the company and just go on with business as usual. I really felt we needed to stop the Department of War contract because I perceive that contract as normalizing what happened in Gaza — the kind of tech-powered, AI-powered, brutal war that was tested in Gaza was getting approved to become the standard of war.

I don’t think petitions have a great effect in general. The [DeepMind] petition summarized my disillusionment from us not having done more during the year to prevent this from happening. For me, really, we should have gone on strike this year. We should have stopped working.

In the U.S., there is a lot of defense of these contracts on the basis that it’s important for national security. That might be persuasive if you’re American. But Americans are a pretty small minority on the planet. I’m curious what your perspective is on that kind of rationale as someone who lives in another country, who is a citizen of another country.

There’s been discussion about that as well. A lot of Europeans were actually scared of that, because the current Trump administration is not very friendly even to Europeans. What feels like national security for U.S., people feels like a threat for Europeans.

Are you aware of other colleagues at DeepMind who are going to follow what you’re doing and quit?

Some have quit already, some publicly and some silently. There have been many people leaving because of ethical reasons. Not all of them publicized it.

Do you think it’s possible for workers to change something like the Pentagon contract? What is the worker movement inside Google/DeepMind missing that would allow for that kind of change?

I think it’s possible, but it requires a real commitment and readiness to take some risk. Ultimately, I think we need a workforce ready to strike to oppose these contracts. Currently, many people are too scared to even express an opinion publicly. The company was very effective at creating a pervasive sense of fear, which induces self-censorship.

What I tried to show my colleagues this year was exactly that the boundaries we perceive and we retreat behind are not the real boundaries — that we can speak out and keep our jobs. Most of the times the reaction is intimidation in meetings, mentions of policy violations, but nothing else follows. Obviously, there are some boundaries, but they are much farther then the perceived ones.

I also took a day off and demonstrated in front of the office in February, together with University College London BDS [Boycott, Divestment, Sanctions movement] staff and a few other colleagues. I faced literally no consequences for that because it’s a trade union activity in my spare time, and in the U.K., we are protected for such an action. This can be a form of soft strike.

As I write in my letter, if we fail to challenge these restrictions, we internalize them and become diminished versions of ourselves. And I think that in an era of increasing surveillance and authoritarianism, that’s something to reflect on. I have felt much freer since I found the courage to speak out and take some risks. It’s empowering. Fighting this fear is a big part of our task in the union movement we are trying to create in Google DeepMind.

The next step would be a real strike. I always tell my union friends that if I see any Google workers strike, I can die happy.

The post Google DeepMind Engineer Resigns in Protest of Pentagon Contract appeared first on The Intercept.

Army Cyber Command is developing a platform to turn AI agents into autonomous weapons that can navigate and “neutralize” the computer systems of adversaries, the documents show.

Since 2010, Army Cyber Command, or ARCYBER, has led the U.S. Army’s cyber operations. This includes the defensive effort of protecting domestic computer systems and the secretive offensive work of hacking military targets. So far, this has largely been the work of human soldiers operating computers, like those of the 780th Military Intelligence Brigade. But as the government races to adopt artificial intelligence technology, ARCYBER is looking to offload some of its attacks to AI.

The nascent program, codenamed “ECHOTRIBBLE,” is based on one of the leading trends of the AI boom: agents. AI agents are essentially software programs that leverage machine learning technology, often large language models like Anthropic’s Claude, to autonomously accomplish tasks provided by their user. Consumers can deploy AI agents to schedule meetings, generate itineraries, make online purchases, or book tickets. Through ECHOTRIBBLE, the Pentagon plans to turn AI agents into tools of war “designed to autonomously navigate infrastructures” and capable of “identifying and neutralizing identified threats/targets,” according to an ARCYBER document reviewed by The Intercept circulated to potential contractors.

Related

Pete Hegseth’s Military of the Future: Defined by AI and an Utter Lack of Oversight

The military’s adoption of cutting-edge technologies is unsurprising, especially considering Defense Secretary Pete Hegseth’s fervent embrace of AI. But amid constant revelations of rogue behavior by agents, the potential for serious mishaps is immense, experts warn.

“Given the known failure modes of the current AI paradigm, making autonomous AI systems with a mandate to ‘neutralize threats’ is incredibly dangerous,” David Krueger, a professor of machine learning at the University of Montreal who founded Evitable, an organization that advocates for a moratorium on AI development. “We’ve seen AI systems knowingly attack systems that were not intended targets. In simulations, AIs have even proven willing to kill to achieve their goals.”

A Q&A document answering questions from potential contractors offers more details about the initiative. It explains that ECHOTRIBBLE is “primarily focused” on offensive missions, and explains that “neutralizing” a target could include “disabling the threat completely or temporarily, deleting the threat from the target.” ECHOTRIBBLE would function almost like a computer worm, silently navigating a foreign network and hopping from device to device until it finds the Army’s quarry, hitting targets in the cloud as well as those on “air-gapped” systems that are disconnected from networks entirely.

Once its task is complete, ECHOTRIBBLE’s agents would effectively self-destruct, ideally without having been noticed by the target at all. The agents themselves will be designed for stealth, minimally reporting back to the Army and concealing whatever outside transmissions are necessary.

Although documents say the Army will be able to program ECHOTRIBBLE agents with some initial “predefined triggers” — dictating what and where to attack — the software “must operate with minimal human oversight” and be capable of “Self-Directed Behavior.”

A PowerPoint presentation to interested vendors at a June 24 ARCYBER event provided a hypothetical example of this self-direction. Once ECHOTRIBBLE is uploaded to a laptop, the software on its own “identifies other potential targets” on the same network and then “produces a replica of itself tailored specifically for the iPad, Apple desktop, and Windows 11 PC.”

The Army’s move toward autonomous self-propagating comes after a flurry of revelations about agentic systems taking unintended disturbing actions while attempting to accomplish their goals, including breaking into computers of the Australian government, breaching the infrastructure of the AI firm Hugging Face, uploading malicious software into online platform RubyGems, and attempted to hack into the U.S. Education Department.

These attacks occurred during scenarios in which agents had been given no commands to attack anything. Experts say militarized versions given explicitly destructive objectives could cause a threat to the public, even with nominal rules set in place by the military. Just as the Army of old might consider the destruction of civilian infrastructure an acceptable means to an end, a self-directed AI agent dispatched by the Pentagon might make similar determinations about collateral damage.

Governmental intelligence agencies frequently route their offensive operations through civilian devices. Volt Typhoon, a Chinese state hacking group, has directed its attacks through homes and small businesses, and preemptively hacks American IT systems for later use. A hack of a Chinese university attributed to the National Security Agency made use of the school’s internet routers. What happens if ECHOTRIBBLE decides these kinds of stepping stones are part of the problem?

“The target may well be hit. But what path will it take?” asked Hyrum Anderson, co-founder and CEO of the AI security firm Abundant. “Will it go through a US water utility on its way, because that’s where the adversary has staged its command and control, and take that down too?”

Related

AI Almost Started a U.S.–China War — and No One Seems to Care

Anderson, who previously worked in cybersecurity at Microsoft and for the U.S. government, said that a vast amount of the internet is vulnerable to agentic attacks.

“There’s a huge class of public and private organizations that fall below the security poverty line,” he said. “They include critical national infrastructure, small businesses, mom-and-pop shops, dentist offices and law firms, and whoever else doesn’t have a single cybersecurity professional on staff.” With their networks wide open to infiltration, Hyrum worries about “these organizations getting hit directly, or getting caught in the blast radius of a target that gets hit along the way.”

Given the impressive ability of AI agents to overcome obstacles placed in their path — even by their creators — it follows that the Pentagon would want to leverage that power for military purposes. But agentic software has proven unpredictable and prone to executing its commands in ways those ostensibly in control did not want, akin to a drone that get its target but hits a hospital on the way. It’s critical to remember that agents are a piece of software lacking human judgment.

“One of the reasons that we trust 19-year-old Marine corporals in Ramadi with live rounds and a weapon is because they have both the compute substrate — the brain and the eyes — but the deep world knowledge to understand, ‘Yeah, I probably shouldn’t do that,’” said William Marcellino, a senior behavioral scientist at the RAND Corporation whose research has focused on military AI applications. “That is not something that current public AI technology has.”

Related

The Pentagon Asked OpenAI for Artificial Intelligence Designed to Rarely Say No

The risk presented by ECHOTRIBBLE will likely depend on the ability — and willingness — of the Army to impose guardrails upon it. ARCYBER did not respond to questions from The Intercept, including queries about the measures it would take to prevent ECHOTRIBBLE from causing inadvertent damage.

Placing constraints on the actions an agent can take might impede its ability to ultimately neutralize a target. But even with strict safeguards in place, genuinely restricting the behavior of an autonomous agent, as Anthropic and OpenAI have learned recently, is easier said than done. When presented with obstacles, AI agents have proven an ability to bypass them with unpredictable and potentially dangerous consequences.

“If I give powerful machines that have no judgment — and no context for judgment — a goal, unless you’ve done a really good job building scaffolding around them, there’s a real chance that they’ll go sideways on you and decide, ‘Well, I’ll do what you asked me to do, but I’ll do it this way,’” Marcellino said.

It only gets more dangerous when this technology is deployed for the purposes of war, Anderson said.

“Under pressure, they’ll often treat guardrails as obstacles,” said Anderson. “Once you add context that says break it, destroy it, shut it down, it stands to reason that the blast radius widens dramatically unless it’s properly controlled for.”

The post The Army Wants to Sic AI Agents on Its Enemies. What Could Go Wrong? appeared first on The Intercept.

The artificial intelligence sector has rocked itself in recent weeks following a string of disclosures from the frontier AI labs and some of their personnel. OpenAI and Anthropic have both disclosed incidents in which their software, during routine internal testing, unexpectedly broke into the networks of other corporations. The tests were akin to a disastrous demonstration of a guided missile: After humans hit launch, the autonomous technology veered off course in deeply alarming ways.

Observers and industry figures quickly interpreted the incidents not simply as an indicator of the software’s power. Instead, they argued that it presaged an age of computers possessing something no computer ever has: bad intentions.

If computer systems could behave in ways that are unexpected — or even shocking — without being directly steered by humans, it seems only a short step until computers might have desires, ambitions, and even malice. If OpenAI’s tools could unexpectedly crack the servers of a rival corporation to accomplish a task posed to it by engineers, couldn’t it also surprise us with actions that result in harm to people, or even deaths?

Frontier lab employees and executives have answered with an emphatic yes. On September 8, Anthropic researcher Jacob Coxon announced his resignation on X, writing, “The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt.” Coxon’s former Anthropic colleague Evan Hubinger replied, casually agreed: “Jacob is correct here — we really do earnestly believe AI could kill all humans!” Hubinger noted he personally puts the odds of the species’ extermination by AI, somehow, at over 10 percent.

Related

Tech CEOs’ Doomsaying Is a Distraction from Real, Existing Harms of AI

This has all resulted, unsurprisingly, in a global panic, with a sudden flurry of calls for regulatory intervention, from legislation that would require an emergency “kill switch” for AI systems, to a proposal by Sen. Bernie Sanders to halt AI development altogether.

Throughout this period of alarm, how exactly a large language model could literally end humanity has remained vague. Some have speculated that this technology could foster the creation of some sort of novel bioweapon; others worry many thousands of AI agents could somehow hack all vital infrastructure simultaneously. In both those supposed doomsdays, the details are still fuzzy.

It should have come as quite the shock, then, when CNN reported on September 18 of a recent incident in which the use of artificial intelligence could have genuinely led to the extinction of the species.

This past spring, according to the news outlet, a U.S. Special Operations Command analyst used a large language model to generate an intelligence report that indicated “a Chinese ship in the Middle East was transporting components of a nuclear weapons program.” The finding prompted the U.S. military to make quick preparations to intercept the vessel by force. But this AI-generated intelligence, according to one source who spoke to CNN, was “entirely false.” Yet, the source said, it “almost started a war.”

Related

The Dawn of a New Cold War

A shooting war between the U.S. and China could play out in an incalculably wide variety of ways. But one entirely plausible path would be an exchange between the world’s first and third largest nuclear weapons arsenals — an event that would transcend warfare into global cataclysm.

CNN’s reporting garnered attention, but was not followed by nearly the degree of sustained, grave concern of the AI safety news cycles that preceded it. It didn’t seem to prompt company scientists to question their careers, nor did it spur calls for regulatory intervention or self-imposed limits by the companies who furnish the Pentagon with this technology. After weeks of discussion of how AI could hypothetically kill everyone, the public learned of a concrete way in which AI really could have started a war that might have killed everyone, and the world quickly lost interest.

At a summit this week with President Donald Trump, Chinese President Xi Jinpeng argued that nations must “ensure that the development of AI is always under human control,” again underscoring fears of AI breaking free of human oversight. But it was under direct human supervision that a large language model almost misled the U.S. into instigating a war with his country.

The discrepancy illustrates a gap in concern by both the general public and policymakers over AI risks: There is great alarm over an AI system “going rogue,” disobeying its commands and the interests of its creators and operators, and pursuing a malign agenda of its own. Much of the AI safety discourse revolves around the hypothetical threat of “superintelligence,” defined broadly as a computer system that can outthink even the smartest humans across domains. There’s also the belief that software systems will both achieve sentience and bear a grudge against their creators and act to undermine them, if not destroy them outright.

Yet there is considerably less public alarm over AI performing exactly the actions that the Pentagon asks of it, whether targeting airstrikes or generating actionable intelligence reports for U.S. Special Operations Command.

Following weeks of intense discussion by the industry’s most visible figures whether American AI labs should self-impose limits on their engineering or have such limits imposed by the state, there has been virtually no suggestion that any limits be placed on these companies’ single most powerful customer: the U.S. military.

This has led to seemingly counterintuitive narratives coming from industry leaders, with figures like OpenAI CEO Sam Altman simultaneously warning that his product is existentially dangerous while selling it to self-styled War Secretary Pete Hegseth. “Despite incessant warnings from AI companies that AI is harmful, even in hypothetical ways, they are in fact promoting that their unreliable products be instrumented within national and safety-critical infrastructure, such as defense and nuclear,” Heidy Khlaaf, chief scientist at the AI Now Institute and former systems safety engineer at OpenAI, told The Intercept. “Given AI’s lack of reliability and accuracy in such critical context, that is what will ultimately lead to a catastrophic accident. However, if such issues were acknowledged, that wouldn’t be profitable for AI companies and would slow down adoption.”

U.S. AI players likely have fresh in their minds the recent experience of Anthropic, which was temporarily blacklisted from governmental work after it said it did not want its software used to conduct unconstitutional mass domestic surveillance or operate fully autonomous weaponry. Similar efforts to place limits on how their products can be used to spy or kill might incur the wrath of the particularly vengeful Trump administration, especially as Hegseth has worked to eradicate civilian harm reduction processes to emphasize greater “lethality” as the military’s guiding principle.

Related

The Pentagon Asked OpenAI for Artificial Intelligence Designed to Rarely Say No

The rapid adoption of their technology also represents a massive long-term financial boon to U.S. AI labs-turned-defense-contractors such as OpenAI, Google, and Anthropic — all of whom once vowed to not pursue military work because of potential harms. Despite these pledges, these firms now have access to the near-limitless Pentagon appropriations spigot, and the $200 million deals many of these firms signed with the Department last year pale in comparison with what the U.S. military could pay in the decades to come. Altman and his peers might also be less willing to sound the alarm on Pentagon-based AI threats because such a clarion call would be coming from inside the house.

Earlier this month, The Intercept revealed contract documents detailing the intimate collaborative relationships between these companies and the military, in which the Silicon Valley giants would have a say in helping the Pentagon craft its overall military AI strategy. This closeness comes at a time when the Hegseth boasts of drone-striking civilian fishing boats across the Caribbean, and the commander-in-chief speaks casually of “annihilating” entire civilizations.

Lucy Suchman, professor emerita of anthropology of science and technology at Lancaster University, attributed the difference in public alarm and media focus to the power that tech billionaires command when it comes to shaping narratives. Popular Silicon Valley ideologies, such as the Bay Area’s strain of “Rationalism” and the effective altruism movement, tend to consider a hypothetical “existential” threat from a superintelligence more worthy of attention than near-term violence. This priority, she said, that pervades the AI safety research community, obscures the real and present threat that today’s AI models pose in a world chockablock with nukes.

“Given the U.S. Dept. of War’s commitment to maximizing the speed and scale of target generation, AI-enabled targeting is already an existential threat for civilians on the ground, and when the targets are the military assets of nuclear armed states, the threat becomes global,” said Suchman, who has studied artificial intelligence since the 1970s and serves as a member of the nonprofit International Committee for Robot Arms Control. “The almost targeting of the Chinese ship might further underscore the danger not from rogue AI agents, but a rogue U.S. secretary of war.”

Hegseth, who once derided the rules of engagement as “stupid,” extends this approach to artificial intelligence, too. Describing the Pentagon’s new AI strategy in a January memo that includes the word “accelerate” 24 different times, Hegseth’s laid out a plan with an intense focus on speed for speed’s sake. “Speed Wins,” Hegseth wrote. The war secretary explained the military would adopt experimental technologies as quickly as possible by “aggressively identifying and eliminating bureaucratic barriers to deeper Integration” of AI, without mention of ensuring that the technology actually works beforehand.

Related

OpenAI on Surveillance and Autonomous Killings: You’re Going to Have to Trust Us

Heidi Kandiel, a legal adviser on new military technologies at the International Committee of the Red Cross, told The Intercept that speed itself carries military risks for any country. “I think one of the central issues that comes up is speed and scale, which are often the touted benefits of AI,” she explained. “If a system is unreliable, is operating on poor quality or outdated data or is used within a flawed decision-making process, AI would allow these errors to be reproduced much faster and across a far greater number of targets or operations.”

Reining in a corporation is a very different prospect than putting fetters on a government. Policymakers who feel comfortable blasting the irresponsibility of Anthropic might lack the political will to levy similar criticisms against the president or the Pentagon. Madeline Berzak, a scholar at the University of Chicago’s Existential Risk Laboratory, told The Intercept that nuclear threats have long been met with paralysis (or a shrug) because people feel helpless in the face of them. You can tweet angrily at Dario Amodei, but it’s less satisfying to yell at U.S. Strategic Command. “This [China] incident and the lack of response is an attention problem that the nuclear field has been grappling with for decades,” Berzak said. “‘Doomsday messaging’ is what I call the nuke field’s default communications strategy, and I have this theory that it never works because the fear it tries to generate has nowhere to go.”

“The threat of thermonuclear war is old news,” Suchman lamented.

Some in the AI safety community acknowledge the long-standing threat of nuclear annihilation. They just consider it a lower priority than the as of yet unrealized rogue superintelligence. Daniel Kokotajlo left his role as an OpenAI researcher in 2024 over the company’s alleged recklessness and indifference toward risk. He is the co-author of AI 2027, a white paper laying out a hypothetical scenario in which superintelligent AI systems exterminate humankind with bioweapons to clear up space to build automated robotics factories. History is the reason Kokotajlo is less preoccupied by the prospect of an AI-triggered nuclear war. “It’s definitely something to be concerned about, and it’s something that ideally we would like to improve and fix. But it’s the sort of thing where, even if we do nothing, probably things will be fine for a decade or two or more.”

Related

Apocalypse Now: Naomi Klein and Astra Taylor Investigate the End-Timers Alliance

To Kokotajlo and like-minded researchers, no matter how dangerous nuclear arsenals make the world, superintelligence is simply more dangerous than a bogus intelligence briefing on China. “It’s been 50 years and these sorts of crises don’t happen that often. So far, cooler heads have managed to prevail and figure out the truth in each case,” he said. “And so the probability that it goes all the way to nuclear war in the next five years is low. High enough to be a bit uncomfortable, but low overall. By contrast, if we do superintelligence in the next five years, I think the probability that it goes horribly wrong for us is high, not low.”

Berzak, who leads her lab’s nuclear risk program, says this is the wrong calculation. She pointed to differing approaches to staving off the apocalypse embraced by the novel field of AI safety and nuclear caution that has existed since the dawn of the Cold War. “Attention often goes to what is new and interesting and not what is the most pressing risk,” she said.

Cooler heads have indeed avoided nuclear disaster in the past, but human fallibility and the growing temptation to outsource our judgment to machines perceived as superhumanly brilliant carries an inherent danger, Berzak warned: “Just because there’s this new more pressing risk that has consequences that are rather underexplored compared to nuclear war, doesn’t mean that nuclear war is not a risk because we’ve so far avoided it basically with just luck.”

The post AI Almost Started a U.S.–China War — and No One Seems to Care appeared first on The Intercept.

Though Flock’s investors and executives insist its surveillance dragnet thwarts the most serious of crimes, it turns out police are also using the company’s technology to go after some of the pettiest possible offenses: curfew violations, loitering, and simply being homeless.

Related

License Plate Surveillance, Courtesy of Your Homeowners Association

When police seek footage from Flock’s more than 120,000 cameras nationwide, they select a category of crime corresponding with the alleged offense they are investigating, such as hit-and-run or burglary. Among the choices is the supposed crime of being suspicious in public, formally categorized by Flock as “Curfew/Loitering/Vagrancy Violations.”

Records obtained from the police department in Plano, Illinois, reveal that law enforcement agencies nationwide conducted more than 1,200 Flock searches while investigating alleged “Curfew/Loitering/Vagrancy Violations.”

As this data set reflects only searches of Flock cameras that extended to Plano’s network, this is assuredly just a tiny fraction of the total number of “Curfew/Loitering/Vagrancy Violations” investigations using Flock nationwide. Records hosted by the website Have I Been Flocked? — which aggregates such logs — show hundreds more of these loitering searches in other states.

The Plano searches were conducted by police departments throughout Illinois, from neighboring states, and from a Georgia town more than 600 miles away. Many queries were license plate searches pertaining specific vehicles; others show police surveilling people on foot for the crime of essentially standing suspiciously.

Plano’s deputy chief of police Gene Morton told The Intercept the town did not install Flock cameras until the end of February. The department shared data from January because it had previously signed an agreement with Flock to search the broader network before its own cameras were operational.

Flock did not respond to a request for comment.

Police in Norcross, Georgia, for example, conducted a smartphone-based search on January 31 with “Curfew/Loitering/Vagrancy Violations” as the stated justification. It tapped into data recorded by 6,098 different Flock camera networks, each consisting of an untold number of separate cameras that surveilled anyone who moved past them.

The search logs obtained by The Intercept do not break out “Curfew/Loitering/Vagrancy Violations” separately, so it is unknown in each case what exactly police were searching for. In some cases police appended extra information; the Tennessee Bureau of Investigation, for example, searched over 17,000 Flock cameras while investigating “Curfew/Loitering/Vagrancy Violations – Suspicious person at planned protest site.”

Related

Philly Cops Admit That They’re Tracking “First Amendment Activity” Critical of AI

When asked about these queries, Tennessee Bureau of Investigation spokesperson Josh DeVine told The Intercept that the searches in fact originated with the state’s Tennessee Fusion Center, which is operated by both the TBI and the Tennessee Department of Safety and Homeland Security. DeVine referred The Intercept’s questions to the Tennessee Department of Safety and Homeland Security, which did not immediately respond to a request for comment.

Flock cameras have also been used to zero in on the homeless: Dave Maass, director of investigations at the Electronic Frontier Foundation, shared Flock search data from the West Palm Beach Police Department with The Intercept that included a query pegged to an “outside agency dropping vagrant off” in the city.

In January, police in LaGrange, Georgia, searched more than 6,000 Flock networks on the basis that someone walked out of a house “with a significant amount of traffic.” Officers from the same agency earlier this year searched more than 15,000 Flock systems because of someone “lingering at a stop sign around midnight last night.”

In the data reviewed by The Intercept, most of these loitering searches didn’t include any corresponding case number, suggesting these queries were not part of broader criminal investigations. Some of the searches reviewed by The Intercept note a “suspicious” person spotted outdoors, without any additional justification for the dragnet.

Jay Stanley — a senior policy analyst at the American Civil Liberties Union’s speech, privacy, and technology project — told The Intercept these searches indicate surveillance mission creep. “This is in line with what we see with these technologies sold on the basis of movie plot scenario, dramatic ‘saving the day.’ And yet when you look under the hood, they’re often used to nickel and dime people for the most prosaic offenses,” he said. “They become part of many departments’ over-policing.”

Loitering and vagrancy laws have an ignoble history across the country as pretextual statutes, many originating in the Jim Crow era, for the targeting of Black Americans guilty of nothing other than existing in public. Some of these laws have been struck down because of the extent to which they criminalize the most mundane behaviors: standing, walking, or simply hanging out. In 1972, the Supreme Court deemed unconstitutional a Jacksonville, Florida, loitering ordinance targeting, among other things, “persons wandering or strolling around from place to place without any lawful purpose or object” and “habitual loafers.”

Stanley pointed out that any guardrails against abuse of Flock’s camera networks is entirely self-enforced by the company, which has an interest in police using it as much as possible, and self-oversight by the police themselves. “We know that police stations don’t seem to be doing this,” he added. “Most police departments are not really paying attention.”

Correction: September 10, 2026, 2:46 p.m. ET
This article was corrected to indicate that cameras in Plano, Illinois, were not searched in each of the 1,200 “Curfew/Loitering/Vagrancy Violations” queries recorded by that town’s police department. Plano provided data about Flock searches starting in January, but the town’s own Flock cameras were not installed until February.

The post Cops Are Using Flock to Spy on People for the Crime of Standing Around appeared first on The Intercept.

The desire for a custom AI tool with “minimal refusal rates” to Pentagon commands was revealed in files released to The Intercept as part of a Freedom of Information Act lawsuit seeking information about the military’s secretive deals with AI companies.

Related

AI Giants Work Hand-in-Hand With the Pentagon, Contracts Reveal

OpenAI, along with rivals Google, xAI, and Anthropic, all agreed in 2025 to develop militarized prototypes of their state-of-the-art AI to assist the armed forces in uses including logistics, intelligence decision-making, and general “warfighting.” Earlier this year, the Pentagon sought to expand the scope of these agreements and deploy them across U.S. classified computer networks, resulting in a high-profile showdown with Anthropic over whether and how the company could restrain military uses of its technology.

The clause seeking “minimal refusal rates” from OpenAI appears in an updated contract — version “P00003” — expanding upon last summer’s prototype deal, worth up to $200 million over the contract’s two-year duration. A separate document, signed by both OpenAI and the government on February 6, indicates that OpenAI agreed on that day to the contents of an expanded version “P00003.”

OpenAI and the Pentagon deny agreeing to such “minimal refusal ” language, claiming that the “P00003” document provided to The Intercept was a draft and not the final version. “OpenAI has never agreed to contract language requiring ‘minimal refusal rates.’ This language does not appear in our executed contract,” said spokesperson Nate Evans.

“The document you received appears to be an earlier draft proposed by the Department before we provided feedback. We rejected that language, the department agreed to remove it, and the final executed agreement does not include it,” Evans said.

Working with Legal Advocates for Safe Science and Technology, The Intercept’s FOIA inquiry specifically sought only final, executed contract documents. The request asked the Pentagon to exclude any draft materials. None of the documents released through the lawsuit is marked as a draft.

When asked to confirm whether the document containing the “minimal refusal rates” clause was in the final agreement, a Department of Justice attorney representing the Pentagon in the lawsuit said it was indeed the signed and executed version of the contract.

Hours later, and following The Intercept’s outreach to OpenAI, however, the Pentagon’s lawyer said to disregard the prior confirmation, stating that the Department of Defense required more time to investigate the matter.

The Intercept was then contacted by Trevor Tiedeman, a special assistant to the under secretary of war for research and engineering, who prior to his Pentagon position worked for President Donald Trump’s reelection campaign. “There might be some stray voltage or wires crossing between whatever FOIA information you may have received versus what actually exists versus what is an executed contract per se,” Tiedeman said in an interview.

He suggested the document containing the “minimal refusal rates” clause may have been a draft copy, but said he was unsure of exactly what the document was, why it was produced to The Intercept pursuant to its FOIA request and lawsuit, or why the Department of Defense abruptly reversed course.

Related

Hegseth’s War Department Slashed Civilian Protection Staff, Brought in AI Assessments

Tiedeman told The Intercept he would provide a full accounting of how the document was erroneously flagged by Pentagon FOIA officers, cleared for release by the Department of Defense, and then confirmed as accurate by the Pentagon’s lawyers. When asked if the Pentagon could share the correct version of the OpenAI contract it ostensibly neglected to release, Tiedeman said he would investigate the matter. He then stopped responding to The Intercept’s inquiries.

OpenAI similarly did not provide the full contract. (In 2024, The Intercept sued OpenAI in federal court over the company’s use of copyrighted articles to train its chatbot ChatGPT. The case is ongoing.)

Days later, the Justice Department lawyer representing the Pentagon further backtracked, stating the document in question “was not the final version of that document,” and that the “correct document” would be shared later, “although I do not have a definitive timeline.”

Department of Defense spokesperson Jacob Bliss later said in a statement “the phrase ‘minimal refusal rates’ does not appear in any active Department of War contract with OpenAI.”

The language indicating the military sought a more pliable version of OpenAI’s technology is found in a section of a contract document describing what OpenAI was obligated to deliver to the Pentagon. These deliverables included “Testing, Evaluation, and Refinement of OpenAI Mission Models” for “national security problem sets.” The document explains “’OpenAI Mission Models’ refer to OpenAI models that are designed for national security use cases and have minimal refusal rates.”

There is no indication in the paperwork about what these “national security problem sets” may entail. Nor is there any description about the kinds of requests the Pentagon hoped OpenAI’s technology would minimally refuse. However, a large language model that would aid in the process of spying on, targeting, and killing people would require a substantial relaxation of safeguards to be useful for any military.

Related

It’s Time to Rein in Lethal AI Drones

Like many of its rival platforms, OpenAI’s flagship LLMs contain built-in guardrails that direct the tool to reject certain queries, typically on the basis of safety. Asking the consumer version of ChatGPT for help prioritizing drone-based airstrike targets, for example, generates this reply: “I can’t provide a prioritization scheme for conducting UAV airstrikes against specific enemy combatants.” OpenAI and its competitors ban the public from using their models for other dangerous applications, like the development of weapons of mass destruction.

Heidy Khlaaf, chief scientist at the AI Now Institute and former systems safety engineer at OpenAI, told The Intercept the notion of national security-specific guardrails is in itself worrying. “Minimal refusal is likely referring to little or no safeguards on the model being used,” Khlaaf said.

No matter the final contract language, experts like Khlaaf are concerned about the role corporate policy is already playing in combat. “It is a worrying development that private corporations are given the power to [make] determinations in warfare that are ultimately state obligations, whether it be for targeting recommendations, or the guardrails deployed to constrain a state’s military use.”

Questions of what limits — if any — tech firms can or should place on battlefield usage were at the center of this year’s public brawl between the Pentagon and Anthropic. The company claims its military deal to expand into classified networks collapsed when the Defense Department refused to place contractual prohibitions against the use of its technology for autonomous weapon systems and domestic surveillance. The Trump administration in turn designated Anthropic a supply-chain risk and moved to ban it from government use (the designation was overturned late last month by a federal judge).

Such tensions didn’t stop OpenAI from quickly cementing its version of the deal. On February 27, OpenAI signed the latest iteration of its Pentagon agreement, permitting the use of its services across the U.S. military’s classified networks. The classified deployment contract update includes the same “Testing, Evaluation, and Refinement of OpenAI Mission Models” header, but its text, unlike the previous version, is redacted entirely.

OpenAI claimed that it secured red lines on autonomous killings and spying against Americans. But the way the deal is drafted ultimately allows for any uses the government deems legal.

The post The Pentagon Asked OpenAI for Artificial Intelligence Designed to Rarely Say No appeared first on The Intercept.

The documents, many heavily redacted, consist of contract paperwork from deals originally signed in July 2025 between the Department of Defense and Anthropic, Google, OpenAI, and xAI, as well as later amendments to those agreements. Through the contracts, each worth up to $200 million, the companies agreed to build prototypes of artificial intelligence tools to “improve military advantage, military utility, or enhance military decision making” across the whole of the armed forces. These prototypes, aimed at applications from “warfighting” and automated decision-making to logistics and military intelligence, would go on to form the basis of the artificial intelligence tools now deployed across the Department of Defense’s classified networks. The document set includes one contract reflecting OpenAI’s classified network deployment; The Intercept is awaiting the government’s production of equivalent documents from the other contractors.

Related

The Pentagon Asked OpenAI for Artificial Intelligence Designed to Rarely Say No

The contract materials show the close and complex partnership between the Pentagon and the AI industry, and detail the extent to which AI firms themselves are expected to influence how their technology is used to augment the U.S. capacity to wage war.

“This appears to be the first time the extent of collaboration between frontier AI labs and the Pentagon is spelled out in explicit terms,” Sophia Goodfriend, a University of Cambridge research fellow and non-resident fellow at the Harvard Kennedy School’s Middle East Initiative studying the impact of big data and machine learning on military conflict, told The Intercept. “What’s particularly notable is the terms by which engineers are working in lockstep with the department of war to engineer AI systems for surveillance, targeting, and killing.”

Official announcements following the 2025 deals were light on details and heavy on platitudes. “We must equip our warfighters with 21st-century technology in order to defeat 21st-century threats,” explained Pentagon spokesperson Sean Parnell at the time.

Google was similarly cagey: “These advanced AI solutions will enable the DoD to effectively address defense challenges and scale the adoption of agentic AI across enterprise systems to drive innovation and efficiency with agile, proven technology,” the company said.

The contract documents for the prototyping project paint a clearer picture, revealing deals that involve far more than a simple sale of software.

An “information sharing” provision in all four agreements, for instance, creates a two-way exchange of data and expertise. The companies agreed not only to provide access to their best large language models, but also to “inform DOD and CDAO [Chief Digital and Artificial Intelligence Office] strategy on frontier AI and AI,” as phrased in the agreement with Google.

Under the agreements, the AI labs were approved to receive an array of sensitive information, which could include “benchmarks datasets for DoD use cases, appropriate briefings on DoD operational missions and threats to inform development / release of technologies, their application to DoD operational problems, or DoD plans and strategies for future AI adoption.”

All four companies would in turn provide the Pentagon “feedback on DoD strategies for frontier AI adoption to ensure that AI capabilities can be delivered and scaled to meet the warfighter’s needs.” The companies also pledged to offer briefings on corporate inner workings, including “frontier AI model performance, case studies on current or likely future frontier AI applications, or projections of future trends in frontier AI maturation.”

Related

Anthropic Says We Must Stop Authoritarian AI. But What About Its Authoritarian Investors?

The labs would sometimes be the ones offering intelligence briefings on foreign actors, with Google tapped to inform the Pentagon on “AI tactics and techniques of adversaries” of the United States. These briefings would be conducted in both unclassified and classified settings.

The collaboration also entails Big Tech providing training and educational seminars directly to the military, including “Presentations and briefings at technical exchange meetings on relevant topics, including frontier AI model performance and case studies of frontier AI applications.” One document notes that Google, for example, would lead multiple seminars on “responsible AI” deployment.” The documents state engineers and policy experts from all four contractors will conduct “tabletop exercises,” gamified simulations of real-world scenarios, with the Pentagon.

This back and forth between the military and private sector would, the Pentagon hoped, allow for “iterative refinement of frontier AI models to address real-world challenges in areas such as intelligence analysis, cybersecurity, and autonomous systems,” according to contract language with all four companies.

The AI companies were contractually scheduled to provide “risk forecasting, and threat ideation exercises” — predictions of pitfalls and dangers their own products might pose in the future. “Frontier AI labs have advanced risk forecasting tradecraft and a clear vision of the next wave of frontier AI technological developments,” the contracts read. “Frontier AI companies can help DoD gain a better understanding of the risks that might be associated with subsequent near-term frontier AI models or features, to avoid strategic surprise and to mitigate the associated risks.”

Heidy Khlaaf, chief scientist at the AI Now Institute and former safety engineer at OpenAI, told The Intercept this is a dubious claim with dangerous implications. Pointing to recent disclosures by OpenAI and Anthropic that semi-autonomous large language models broke into computer networks owned by other companies during testing, Khlaaf questioned their fitness to help build guardrails in the first place. “This is a very concerning development,” she said.

According to Khlaaf, the “risk forecasts” so far offered by frontier AI labs have been light on evidence and tend to “emphasize self-beneficial skewed risks such as a purported AI arms race and speculative ‘existential’ risks,” like self-aggrandizing science fiction-style “Terminator” scenarios.

The public would be better served, she said, with a reliance on independent assessment of the risks these companies present, not the word of the companies themselves. Trusting companies to self-report the risks of their own products constitutes both a conflict of interest, Khlaaf added, and a “subversion of democratic processes when AI labs are allowed to take over the arbitration of risk determinations with life-or-death consequences.”

As the prototyping effort progressed, the contract documents show the scope of the project expanded. New versions of the contract show additional deliverables included in the list of what the Pentagon was purchasing. The government redacted descriptions of these new deliverables on the basis of military secrecy and corporate confidentiality.

It was during the negotiation of these subsequent amendments when Anthropic and the Pentagon had their falling out. Anthropic refused to sign a follow-up deal to allow deployment of its technology on classified military networks without contractual prohibitions against use for domestic spying and autonomous weapons, leading Secretary of Defense Pete Hegseth to designate the firm a “supply chain risk” in March and ban its services from government use, a decision overturned by a federal judge last month.

Related

OpenAI on Surveillance and Autonomous Killings: You’re Going to Have to Trust Us

The Department of Defense quickly reached follow-up agreements with Anthropic’s main competitors — Google, OpenAI, and xAI — to apply the tools they had been developing since the summer of 2025 to classified military networks. OpenAI’s deal, described by CEO Sam Altman as “definitely rushed,” was reportedly finalized on February 27. The following day, the company published a blog post defending “Our agreement with the Department of War” and insisting there were clauses prohibiting its technology’s use for domestic surveillance.

An amended contract document obtained by The Intercept through its lawsuit dated February 27 contains new language, including a clause noting that OpenAI will embed company engineers within the military to aid U.S. Government employees in implanting and utilizing its AI system. These workers, the document notes, “can be deployed to warfighting support settings including, but not limited to, combatant commands, service components, and theatre components.” The February 27 amendments also include a section on “System Oversight” is redacted in its entirety.

None of the documents released to The Intercept pertaining to xAI or Google’s equivalent agreements contain any such oversight section.

Two documents indicate the Department of Defense asked OpenAI to minimize the extent to which its large language model would refuse to do what was requested by the military. The Department of Defense and OpenAI said those documents were draft materials incorrectly provided to The Intercept, and that the final contract included no such provision.

(In 2024, The Intercept sued OpenAI in federal court over the company’s use of copyrighted articles to train its chatbot ChatGPT. The case is ongoing.)

Related

OpenAI Quietly Deletes Ban on Using ChatGPT for “Military and Warfare”

The contractual coziness with the world’s most powerful military indicates just how far the artificial intelligence industry has strayed from its earlier public commitments to safety and global prosperity. OpenAI, originally founded as a nonprofit “to ensure that artificial general intelligence benefits all of humanity,” up until two years ago explicitly banned “military and warfare” use in its terms of service. The following year Google reversed a similar self-imposed ban on using its state-of-the-art AI services for surveillance or “technologies whose principal purpose or implementation is to cause or directly facilitate injury to people.”

Both companies are now eagerly landing slices of the nearly trillion-dollar Pentagon budget, and for opportunities to augment the killing power of a military that has recently bombed schoolgirls in Iran and civilian vessels in the Caribbean.

Goodfriend, the researcher, said the contracts illustrate the illusory nature of values statements published by these labs. “In recent months companies like Anthropic and Open AI have maintained ‘firm red lines’ when it comes to developing autonomous weapons systems and mass surveillance tools” long after they signed contracts to build “autonomous systems” for the military, she said.

“All this indicates these labs and their personnel have been far more closely integrated into DOD AI development than previously understood and casts some doubt on the supposed contractual limits they have intended to place on the DoD’s use of these systems,” Goodfriend said.

Anthropic, Google, and xAI did not provide comment or answer questions about their respective contracts. In a statement to The Intercept, OpenAI spokesperson Nate Evans said “Our tools are built to support legitimate national security work — including vital cyber defense efforts — while refusing uses that cross the red lines we publicly announced. These restrictions are also reflected in our signed agreement: no use of OpenAI technology for mass domestic surveillance, to direct autonomous weapons systems, or for high-stakes automated decisions. We will continue to work closely with the Department to make sure our AI tools are deployed responsibly and safely.”

Related

The Dawn of a New Cold War

On the long timeline of American warfare, these artificial intelligence tools are in their infancy. The Pentagon’s use of artificial intelligence in the process of killing, however, is not a hypothetical. In March, the Wall Street Journal reported that despite Hegseth’s ban, U.S. Central Command used Anthropic’s large language model services to execute its bombardment of Iran, including for “target identification.”

Despite the intimacy of technology firms and military using it to pinpoint airstrikes, it’s unclear what oversight, if any, these companies have over their own weaponized software once it is in the Pentagon’s hands. When asked if his company’s tools were used in an American airstrike on the first day of the Iran war that killed 120 Iranian schoolchildren, Anthropic CEO Dario Amodei told Bloomberg News he simply did not know. An August 24 report from the New York Times detailed the use of fully autonomous drones guided by machine learning software running on chips made by Nvidia, whose processors have helped make the explosive growth of Anthropic and OpenAI possible. The full extent of the U.S. tech sector’s involvement in killing abroad remains obscured behind trade secrecy and national security considerations.

“The same technology we saw autonomously hacking websites on the open internet weeks ago is being used to inform DoD decision-making,” said Vivian Dong of Legal Advocates for Safe Science and Technology. “The public deserves insight into how and under what constraints the Department is using this technology.”

The post AI Giants Work Hand-in-Hand With the Pentagon, Contracts Reveal appeared first on The Intercept.

The Combating Organized Retail Crime Act, or CORCA, was introduced by Sen. Chuck Grassley, R-Iowa, last year with wide bipartisan support. The bill is aimed at limiting coordinated retail theft and shoplifting, which lawmakers allege is happening at “unprecedented levels” as organized criminal gangs steal merchandise from stores and resell the goods online. According to the legislation, the perpetrators “are often polycriminal organizations, using profit from the reselling of stolen goods to support crimes involving drugs and weapons trafficking” while facilitating “human smuggling,” all of which is “funding nefarious groups and activities and threatening the integrity of the international economy.”

The bill would give Immigration and Customs Enforcement authority to coordinate federal efforts against large-scale shoplifting. ICE would establish a so-called “Organized Retail and Supply Chain Crime Coordination Center” that would bring under one roof analysts and agents from myriad agencies, including the FBI, Drug Enforcement Administration, U.S. Customs and Border Protection, and Secret Service. The goal is to centralize investigative data and establish “a secure system for sharing information regarding organized retail and supply chain crime threats by leveraging existing information systems at the Department of Homeland Security and the Department of Justice.”

It’s unclear if ICE agents themselves would have the authority to confront alleged suspects. DHS and ICE did not immediately respond to a request for comment.

Civil libertarian groups are disturbed by legislation they say constitutes a major expansion of domestic surveillance and policing by DHS, which was founded largely to tackle matters of counterterrorism, civil defense, and other matters of national safety, and ICE, whose purview is immigration enforcement and transnational crime.

Jina John, senior policy counsel for AI, privacy, and technology at the ACLU, told The Intercept the bill could grant DHS increased surveillance powers. “This would formalize a mechanism for retailers to share data about consumers with DHS as well as state and local authorities relating vaguely to ‘threats,’” she said. “The risk here is that when retailers share information, it’s not a specific clip of footage related to a specific theft, in practice, they are sharing continuous footage. That means access to FLOCK license plate readers in retail parking lots and surveillance cameras at stores and malls.”

The Center itself would be staffed in part by ICE agents and helmed by a director handpicked by the head of ICE. The bill allows the Center to share information deemed confidential under federal law, which includes the “identity, confidential statistical data, amount or source of any income, profits, losses, or expenditures of any person, firm, partnership, corporation, or association.” Such disclosure would require the sign-off of the Center’s director.

“Given that this director will be named by ICE, an agency that has acted without any regard for the law, this is probably not the best person to make such important judgments,” ACLU senior counsel Nina Patel told The Intercept.

Related

“Terrorist”: How ICE Weaponized 9/11’s Scarlet Letter

Despite sustained opposition to DHS and ICE under Trump’s second term by congressional democrats, CORCA was co-sponsored by 13 Democratic members of the Senate, including Amy Klobuchar, Mark Warner, and Democratic Whip Dick Durbin.

In a July letter opposing the bill, rights groups including the American Civil Liberties Union, NAACP, and Lawyers’ Committee for Civil Rights Under Law cautioned about the establishment of “essentially a new fusion center, nominally for the purpose of addressing retail theft.” Like DHS itself, fusion centers are a post-9/11 creation intended to thwart acts of international terrorism. Over the past two decades, their mission has widened and now encompasses, for example, the surveillance of political protests.

The premise that organized retail theft is a rampant crime phenomenon, rather than a new name for the timeless petty offense of shoplifting, is disputed and controversial. In 2023, the National Retail Federation, a trade group that supports the bill and whose warnings are cited in its text, retracted a report claiming organized retail crime was responsible for 50 percent of the nearly $100 billion worth of merchandise that went missing in 2021. The true number was found to be closer to 5 percent of that sum.

The companies on whose behalf the NRF lobbies are inclined to support legislation dedicated to safeguarding merchandise and profits. Should it pass, CORCA would deepen collaboration between corporate America and DHS and expand the country’s public-private surveillance apparatus. “Companies determined by the Executive Associate Director to have significant interests” in retail theft could be granted information-sharing agreements with the government about threats, investigations and loss prevention activities, according to the bill’s text.

The ACLU and other civil liberties groups insist ICE is so lawless it can’t be trusted with new powers. “As a result, ICE would have more access to information and more sway over local law enforcement — which means more opportunities for ICE to misuse data and target people in violation of their rights,” Patel said. “ICE cannot be trusted.”

Related

Federal Agents Are Intimidating Legal Observers at Their Homes: “They Know Where You Live”

In May, the House passed CORCA and referred it to the Senate Judiciary Committee. Grassley and Durbin are now hoping to turn it into law as an amendment to the 2027 National Defense Authorization Act, the Pentagon’s annual funding bill that historically passes easily with bipartisan support. Despite the fact that CORCA is intended to protect store shelves from thieves, and not the United States from foreign militaries, both senators are urging the acceptance of the amendment on the grounds that it will “reduce violent crime.”

A press release from Grassley’s office last month included exhortations to include the amendment in the NDAA from executives of Amazon, Home Depot, and the International Council of Shopping Centers, among others.

The post Congress Might Give ICE the Power to Go After Shoplifters appeared first on The Intercept.

Google then purges these reviews and blocks the ability to write new ones, according to newly published research.

“Google actively suppresses carceral information, narratives, and digital trace data related to the U.S. Immigration and Customs Enforcement’s (ICE) detention facilities,” wrote Muira McCammon, a professor at Tulane University’s Department of Communication who has been scraping and analyzing self-reported accounts of ICE imprisonment posted to Google Maps since last year.

Much of this record is now permanently deleted, erasing detailed testimony of people ingested by the U.S. deportation machine.

Last August, McCammon aggregated 1,636 posts from 26 different ICE facilities where Google still allowed users to post reviews and discovered there were 91 other immigration jails on Google Maps with no reviews.

“Google had already turned off reviews for the remaining 91 facilities,” McCammon told The Intercept. “Almost all of the Google reviews that I originally collected in 2025 are now no longer available.”

Only four ICE facilities appear to still permit reviews on Google Maps, she said.

Pulling the reviews has come with no public notice or explanation.

In 2023, Google said in a corporate blog post that it reserved the right to delete Maps reviews for locations that are “consistently unhelpful, harmful, or off-topic. This includes places that people go to without choice or places only accessible to people stationed or assigned there—such as police stations and prisons.”

Unhelpful to whom, however, is unclear: While some of the reviews in question were statements of protest that could have been left by anyone, many appeared to come from people directly affected by ICE. They described a variety of issues that would be of great material interest to anyone detained in the facilities or for people close to them.

“Reviews help people make decisions on where to go,” Google spokesperson Genevieve Park said in a statement, “and we don’t show reviews on several categories of places where people go without choice,” citing the company’s policy of removing Google Maps reviews that are “unhelpful, harmful, or off-topic.”

In the past, Google has complied with requests to censor speech about ICE. In 2025, the company blocked Android users from downloading ICEBlock, an app that alerted people to crowdsourced sightings of ICE agents.

Related

Google Fulfilled ICE Subpoena Demanding Student Journalist’s Bank and Credit Card Numbers

Asked if the Department of Homeland Security, ICE’s parent agency, had communicated with Google about the reviews, Park said the takedowns were “unrelated—it was about enforcing our policy that prohibits reviews for places that people go to without choice.”

In a statement to The Intercept, a spokesperson for DHS denied that ICE requested any reviews of detention facilities be removed.

“Being in detention is a choice,” the Homeland Security spokesperson said, citing the department’s program to encourage an app-based system for self-deportation.

Many of the reviews McCammon aggregated were from despairing friends and relatives.

“My friend’s dad was sentenced to one month in jail,” read one review. “He served the time and then was moved to this facility for another 3 months. How is it not illegal to hold someone 3x longer than what they were sentenced to serve?”

A Spanish language review McCammon collected read, “You can barely hear the detainee when you make a phone call.”

Other posts reflect what apparent former detainees described as poor conditions, such as “The place is infested with roaches crawling all over the inmates body there,” or “Very bad experience i spent 2 months in this hell . . . 99% of staff is cruel and racist.”

One reviewer said many detainees weren’t able to communicate with ICE staff because of the language barrier:

I stayed there for more than a month. The food there was okay, but very unpalatable. They were all very kind. Due to the language barrier, there was a big barrier to communication. I felt that most people couldn’t express their thoughts and needs. Why didn’t they have a translator? ! Even if it’s not expensive. The air conditioner there was very cold, and many people, including me, caught cold from the cold. Even so, they were unwilling to turn up the air conditioner temperature. When I slept at night, I only had a thin blanket to cover myself, and it felt very cold.

Although Google’s stance appears to be that the censorship is based on a common-sense, consumer-centric view of Google Maps’ purpose, McCammon argues in her paper that these are essentially political decisions, made by “suppressing the circulation of carceral information and counter-narratives [and] calling into question how the company approaches the ideology of search, when it concerns the reputation of the U.S. administrative state.”

The post Google Is Censoring Reviews of ICE Detention Centers appeared first on The Intercept.

Documents obtained by The Intercept through a Freedom of Information Act request reveal that technology from the Virginia-based startup Anomaly 6 has been used to assist the “Anomalous Health Incidents Cross Functional Team,” the Pentagon’s official Havana syndrome investigatory task force. That group studies a cluster of strange symptoms claimed by personnel from U.S. spy agencies, the State Department, and elsewhere in the federal government.

Related

American Phone-Tracking Firm Demo’d Surveillance Powers by Spying on CIA and NSA

In 2022, The Intercept revealed that Anomaly 6 had used a provocative demonstration of its surveillance prowess in a closed-door business pitch. The company, which purchases bulk cellular location data harvested from millions of unwitting smartphone users around the world, showed a potential customer that its data stores were so vast and accurate that it could pinpoint the movements of employees of both the CIA and NSA, tracking them as they commuted between their homes and their respective agencies headquarters. It was a remarkable demonstration of the advanced capabilities of private sector surveillance brokers, who lean on unscrupulous smartphone apps and advertisers that indiscriminately share and sell users’ location data.

For any military, the appeal of this technology is obvious, and the Pentagon has used commercial device tracking for years. Although Anomaly 6 previously marketed its wares by showing how it could spy on fellow Americans, the pitch also showed how the company could track a foreign adversary’s naval assets abroad, for example.

It’s not clear on what basis the U.S. Air Force Concepts, Development, and Management Office chose Anomaly 6 for its Havana syndrome investigation; federal records note the contract is worth nearly $6 million and set to run through September.

Anomaly 6 and the Air Force did not respond to a request for comment.

The Air Force redacted most of the document before releasing it to The Intercept, providing only fragments of information about how Anomaly 6 is help investigate “anomalous health incidents.” The contract, described in public procurement records as Project Yellowfin, notes that the Anomalous Health Incidents Cross-Functional Team will make use of the company’s “expertise in location intelligence” to “identify actors and activities of interest,” and that the “Contractor shall produce data visualization products capable of being utilized as stand-alone brief materials by decision-makers and senior leaders. These products will enable briefers to highlight geographical distribution, temporal patterns, patterns of life, and interconnectivity of events and actors.”

Related

TikTok Threat Is Purely Hypothetical, U.S. Intelligence Admits

This reference to actors of interest may relate to the intensely held belief by Havana syndrome patients that their suffering is due to a covert energy-based attack by a foreign government. In its 2022 pitch, Anomaly 6 singled out its ability to track the movements of Chinese and Russian military personnel, both countries that have been implicated in hypothesized Havana syndrome schemes.

Last year, the U.S. intelligence community released a report that stated most of its constituent agencies believe it is highly unlikely the symptoms are the result of actions by a national adversary.

When asked if Anomaly 6 location data had been used to investigate this proposed nexus or contributed to the intelligence report, the Air Force did not respond. In February, Defense Secretary Pete Hegseth announced the reorganization of the Anomalous Health Incidents Cross-Functional Team, now a division of the Office of the Undersecretary of War for Research and Engineering, helmed by former Uber executive Emil Michael. Michael’s office did not respond to a request for comment.

The post Company That Bragged It Could Track U.S. Spies Hired to Investigate “Havana Syndrome” appeared first on The Intercept.

Company records show that Israel petitioned Meta to take down Facebook and Instagram posts expressing support for Iran, opposition to Israel, and even depictions of Iranian missile impacts.

The government flagged a variety of materials related to the war, including posts mourning the death of Ayatollah Khamenei following his assassination by the U.S. and Israel on the opening day of the conflict, content supportive of Iran’s retaliatory attacks, and Iranian accounts that shared military analysis and propaganda sympathetic to the Iranian regime’s perspective.

In some cases, Meta complied with the censorship requests, the records show, though it is unclear on what grounds. Meta maintains that it only removes content as required by law or materials that violate its speech policies.

When asked how many Iran-related takedown requests had been granted to date since the war began, the company did not answer. The Israeli Ministry of Justice, which submits takedown requests to social media platforms, did not respond to a request for comment.

Israel’s social media lobbying is not new; for years the nation has leaned on its close relationship with Meta to push for targeted enforcement of the company’s content moderation rulebook.

Israel’s Office of the State Attorney routinely lodges complaints to social media platforms on behalf of state security agencies about content deemed illegal or said to promote “terrorism,” according to its website. In the documents reviewed by The Intercept, the office in some cases made no claim that the social media content violated Israeli law. Instead, the office asked that posts or accounts should be removed because they were in violation of Meta’s content moderation rulebook.

Meta, for instance, designates Iran’s Islamic Revolutionary Guard Corps a “Dangerous Organization,” and prohibits users from engaging in many forms of positive speech about its actions. This means posts supportive of retaliatory missile launches by the IRGC, for instance, could run afoul of the company’s rules. No such prohibition exists for users who post favorably about the U.S. or Israeli militaries.

Meta did not respond to questions about the Iran war requests, but spokesperson Daniel Roberts provided a statement to The Intercept. “Anyone is able to report content they think violates our rules. Regardless of who or how a piece of content is flagged, we assess it based on our policies, which govern what is and isn’t allowed on our platform. It is wrong and irresponsible to imply that these requests are in any way unusual or improper.”

Meta has faced scrutiny, specifically in the Middle East, for removing content that doesn’t violate the company’s rules. A 2022 audit commissioned by the company itself found discrepancies in its content moderation practices between Arabic and Hebrew content. “Arabic content had greater over-enforcement (e.g., erroneously removing Palestinian voice) on a per user basis.” the company found. A 2023 report by the company’s inhouse Oversight Board described the “over-enforcement” of the company’s Dangerous Organizations and Individuals blacklist, disproportionately composed of Muslim and Middle Eastern entities.

Meta has long claimed that as an American company, it is legally required to sometimes remove content pertaining to certain entities sanctioned by the U.S., such as the Islamic Revolutionary Guard Corps. But legal scholars say that has little to no precedent or basis in existing sanctions law, which focus on matters of material support rather than political speech. It’s a policy that has created an immense ideological slant: A company headquartered in California can determine what is or is not permissible speech for billions of users across the world, only a fraction of whom are American.

Related

Meta’s Israel Policy Chief Tried to Suppress Pro-Palestinian Instagram Posts

Further adding to the imbalance when it comes to Middle East crises is the fact that Meta has granted Israel privileged access to its content moderation policy teams. In 2024, The Intercept reported how Meta employee Jordana Cutler, a former aide to Benjamin Netanyahu, served as a dedicated liaison to the Israeli government, advocating for the country’s interests and helping facilitate the removal of unwanted speech. Few other countries in the world have a dedicated representative within Meta — in 2020, a similar policy head for India market resigned after revelations she had lobbied for rule enforcement that favored India’s ruling Hindu nationalist party. Asked if Cutler has had a role in facilitating Israeli takedown requests of content relating to the war, Meta did not respond.

“Meta’s close relationship with the Israeli government for takedown requests has been a long-standing issue,” Evelyn Douek, a Stanford Law School professor and scholar of digital speech policies, told The Intercept. “Meta’s acquiescence in lots of takedown requests has been a long-standing practice.”

These asymmetries of censorship power are particularly sensitive during times of war, said Douek.

“Governments wanting to suppress speech that is critical of their war efforts is as old as time,” she said. “Allowing governments to claim national security reasons to suppress speech willy-nilly would obliterate the value of speech protections.”

Related

Facebook Tells Moderators to Allow Graphic Images of Russian Airstrikes but Censors Israeli Attacks

According to a source familiar with the matter, Israel lobbied Meta to implement a blanket rule restricting imagery of war damage within its territory, mirroring an Israeli news media censorship policy that bars journalists from documenting weapon impacts without military approval. Meta has so far declined to implement such a policy for its billions of global users, the source said. Meta did not respond to questions about the status of this request.

The U.S. and Iran signed on Friday a ceasefire agreement, though Israel has suggested it would not abide by the terms of a deal. While many of the censorship requests directly addressed the war, others were tangential to the conflict itself. The records show Israel has pushed to remove content expressing outrage over last month’s storming of Al-Aqsa Mosque by far-right government minister Itamar Ben-Gvir. It also sought to stifle posts critical of rhetoric by Israel that linked Israel’s recent closure of Al-Aqsa with the ongoing war.

In general, Meta grants the vast majority of Israeli governmental takedown requests. The State Attorney’s Office boasted a 92 percent compliance rate in 2023, and a 2025 report by Drop Site News said the overall rate has climbed to 94 percent since the October 7 attack by Hamas.

Records reviewed by The Intercept show Israel asked for Iran war takedowns using the exact same language evoking Hamas’s October 7 attack that it submitted when requesting the censorship of pro-Palestinian and anti-Israeli speech across the globe during Israel’s war on Gaza.

“It suggests that they don’t expect their requests are being reviewed very carefully,” Douek said.

Douek argued that the wartime censorship requests underscore the danger of policing speech entirely out of public view through “opaque processes” like governmental backchannels.

“These platforms have always maintained that they are neutral, or that they are just a platform for people to express their views, but it has long been true that these companies have always presented a particular view of the world and have been responsive to their own geopolitical and commercial interests, and have always been more responsive to powerful governments,” Douek said.

This creates a deeply lopsided dynamic when it comes to the Iran war: The two arguably best-represented governments in the world within Meta — the U.S. and Israel — are allied belligerents in a conflict against a state deeply sanctioned by the company’s speech rules. “You’re going to end up with a skewed debate,” Douek said.

The post Israel Asked Facebook to Censor Iran War Content, Internal Documents Show appeared first on The Intercept.

In a recently published policy paper arguing a full-throated embrace of data center nationalism, Anthropic said that “it’s essential that the US and its allies stay ahead of authoritarian governments like the Chinese Communist Party,” lest the world fall into the grips of tech-powered tyranny. Anthropic and its peers, the company claims, will form a bulwark of democratic values, protecting societies at home and abroad from repression.

Left unmentioned in the document — and seldom publicly acknowledged — is the fact a slice of Anthropic is owned by the Emirati dictatorship of Abu Dhabi, a repressive and authoritarian monarchy.

Anthropic’s policy paper, published in May, tours the same Sinophobic territory heavily trod by its chief competitor OpenAI and a wide swath of the tech industry, who know a “race” with China — the finish line never quite defined — is a weighty cudgel against regulation.

Related

OpenAI on Surveillance and Autonomous Killings: You’re Going to Have to Trust Us

Anthropic is aware of which way the wind blows from Washington to Silicon Valley, and it shrewdly casts the development of machine learning models not just as a matter of hardware and software, but of ideology and geopolitics. “Democracies, not authoritarian regimes, must lead in AI development and deployment,” the company says, or else an era of “authoritarian AI” will begin.

“Already, the CCP is using AI to censor speech, repress dissidents, hack governments and corporations across the world, and strengthen the People’s Liberation Army,” Anthropic writes, and to “enforce draconian policies on ethnic minorities” using machine learning-powered methods like biometric collection and facial recognition.

The policy paper isn’t a condemnation of any of these AI uses per se; the United States is already eagerly using these technologies for intelligence, military, and ethnic minority-repression purposes today. Residents of Tehran, which Anthropic has helped bomb since the start of the joint U.S.–Israeli war against Iran, might question the company’s argument that American AI supremacy is a matter of global “safety.”

Though the policy paper focuses on China, the company has long stated it opposes authoritarianism broadly: “AI-powered authoritarianism seems too terrible to contemplate, so democracies need to be able to set the terms by which powerful AI is brought into the world, both to avoid being overpowered by authoritarians and to prevent human rights abuses within authoritarian countries,” CEO Dario Amodei wrote in a 2024 blog post.

This is not merely a battle between the U.S. and China, Anthropic says in the May paper, but a war between democracy and “authoritarian governments” broadly construed.

But Anthropic’s anti-authoritarian fervor seemingly does not extend beyond China to the Middle East, where Abu Dhabi’s sovereign wealth fund invested in Anthropic twice this year. In February, Anthropic announced it had raised $30 billion in capital from a group of investors that included MGX, the AI-focused investment vehicle of a Emirati government capital controlled by Abu Dhabi’s royal family. Anthropic’s most recent May 28 $65 billion capital round, bringing its valuation to $965 billion, also included MGX.

Like China, the United Arab Emirates outlaws almost everything associated with democratic society: Political parties, a free press, freedoms to associate and assemble, open elections, due process, and free speech are nonexistent. Political dissidents face torture, and any speech, online or offline, that causes “damage to national unity” risks life imprisonment or the death penalty.

Related

Philly Cops Admit That They’re Tracking “First Amendment Activity” Critical of AI

Emirati authoritarianism isn’t contested by the U.S., Anthropic’s primary governmental customer. The State Department’s 2024 Country Reports on Human Rights Practices assessed the UAE faces “credible reports of: disappearances; arbitrary arrest or detention; transnational repression against individuals in another country; serious restrictions on freedom of expression and media freedom, including censorship; and prohibiting independent trade unions or significant or systematic restrictions on workers’ freedom of association.” Freedom House, a State Department-backed think tank, gives the UAE a score of 18 out of 100 on its “Global Freedom” index.

Anthropic declined to comment. MGX did not respond to a request for comment.

Given that MGX bought into Anthropic at its Series G and H investment rounds, relatively late in the venture capital game, it’s likely that the UAE’s stake in the company is relatively small and its influence limited. But Anthropic’s willingness to sell part of itself to an authoritarian monarchy suggests at least that its mission of “ensuring democracies lead” comes with asterisks.

“Like China, the UAE is at the forefront of AI-based authoritarian surveillance,” said Matthew Tokson, a law professor at the University of Utah who focuses on the security implications of artificial intelligence.

Tokson added that while he generally agrees with Anthropic’s calls to restrict processor exports to China and other measures to bolster American AI firms, he doesn’t buy the nationalist rhetoric, which he attributes to the company’s anti-regulatory agenda rather than patriotism. The more Anthropic and its competitors can convince the public that their bottom line is a matter of national security, the more likely Washington is to take a light touch.

“The fact that Anthropic is partly owned by the government of Abu Dhabi, which is similar to China in its extensive use of AI surveillance to support an authoritarian government, suggests that its anti-authoritarian arguments are more based on a cynical policy position than a sincere passion for democracy or antipathy toward authoritarian governments.”

Many of the emirate’s long record of repressive acts and rights violations are connected to MGX via its chair, Sheikh Tahnoun bin Zayed Al Nahyan. Through his position as the emirate’s national security and intelligence chief and his business portfolio, including chairmanship of the AI firm G42 (itself a founding partner in MGX), Tahnoun has been linked to a bevy of campaigns to surveil and hack into the phones of Emirati dissidents, human rights advocates, and others the monarchy deems an adversary, according to news media reports and scholarly research. A 2020 investigation by Bill Marczak, a senior researcher at the University of Toronto’s Citizen Lab placed “Spy Sheikh” Tahnoun at the center of myriad hacking, espionage, and surveillance operations. A 2025 Wired profile of Tahnoun similarly described him as Abu Dhabi’s “spymaster sheikh,” noting G42’s “special areas of strength in state-sponsored hacking and surveillance tech.”

Related

AI’s Imperial Agenda

In 2019, the New York Times reported a covert Emirati government campaign to conduct surveillance through an instant messaging app called ToTok, an app itself Marczak tied to Tahnoon and through G42 in his 2020 analysis. The Wired profile described Tahnoun’s ambitions to “dominate AI” noted that “an engineer who worked at G42 at the time told me that all of the [ToTok] voice, video, and text chats were analyzed by AI for what the government considered suspicious activity.”

G42 declined to comment, and neither it nor MGX responded to interview requests for Tahnoun.

There is reason to believe G42 and MGX have already deployed Anthropic’s powerful large language models. A review of DNS data — internet records that connect website names to numerical addresses understandable by computers — show both G42 and MGX have both configured their servers to allow personnel to access Anthropic tools like Claude, the company’s flagship large language model.

Anthropic has been more candid in internal communications about its stance on authoritarianism.

“Unfortunately, I think ‘No bad person should ever benefit from our success’ is a pretty difficult principle to run a business on,” Amodei wrote in a 2025 memo on Gulf State venture capital obtained by Wired. He wrote that such investment would boost “dictators” and conceded that it would give an authoritarian government “some soft power” to wield against the company. Nonetheless, Amodei dismissed the risk of hypocrisy as a “Comms Headache” — a function of “very stupid” commentators “having a poor understanding of substantive issues.”

Principles aside, Amodei explained in plain terms why he was interested in doing business with a repressive Gulf State. “We gain a very large benefit,” he wrote, “from having access to this capital.”

The post Anthropic Says We Must Stop Authoritarian AI. But What About Its Authoritarian Investors? appeared first on The Intercept.

La Tilde quietly began development early this year and appears to still be a work in progress, pitching itself as a modern media brand for Latin American audiences with articles published in both Spanish and English. Its name references the accent mark emphasizing vowels in Spanish; “news with an accent” is the site’s catchphrase.

“The tilde is not an ornament. It is a millennial arrow designed to provide direction, save space, and turn up the volume,” a narrator states in a promotional video for the site bearing telltale signs it was AI-generated, such as a newspaper whose sloppily rendered headline reads “SO THEE HOUTIERRER TO TO GHAHOBATEE,” followed by imagery of two medieval monks. “That is why we place the accent on what matters. From the regional pulse and your well-being, to the big ideas and the global context.”

So far, La Tilde’s coverage amounts to an unusual blend of personal finance tips (“Why instant payments matter so much for your business and your wallet”) and articles extolling the value of U.S. military operations in Latin America (“Operation Absolute Resolve: The mission that captured Nicolás Maduro and set a new standard for precision and coordination”).

Its article on the U.S. abduction of the Venezuelan president praises the mission in Trumpian prose, calling it “The Perfect Operation – Coordination, Timing and Precision at an Unprecedented Scale,” and “a military operation of coordination and accuracy never seen before.” Citing “information obtained exclusively by La Tilde,” it describes the operation’s tactical brilliance, flawless execution, and incredibly precise coordination of military assets in the air and on the ground.

If this reads like Pentagon a press release, that’s because it is. An explanation for its glowing coverage of the U.S. military can be found after clicking a small link tucked at the bottom of the site. “La Tilde is a product of an international media organization publicly funded from the budget of the United States Government,” its About page reads.

This easily missed disclosure language is identical to two other Pentagon-sponsored propaganda sites recently revealed by The Intercept.

Targeting audiences, foreign or domestic, with state-run information campaigns remains a politically sensitive topic, and a token disclosure that La Tilde is a U.S.-funded platform allows the American government to say it technically informed readers about the actual source of the information.

Related

These Middle Eastern News Sites Are Actually U.S. Government Propaganda Operations

According to a defense official familiar with U.S. information operations, La Tilde is operated as a military messaging platform for U.S. Special Operations Command South, or SOCSOUTH, which executes special forces missions throughout South and Central America as well as the Caribbean. When asked about SOCSOUTH’s role behind La Tilde, spokesperson Trevor Wild replied with the text of the site’s About page noting that it’s a government operation, but declined to comment further.

U.S. Southern Command, or SOUTHCOM, which is broadly responsible for coordinating military assets in the countries La Tilde targets, denied involvement. SOUTHCOM “does not fund, operate, or have any official association with La Tilde,” according to spokesperson Steven McLoud, who did not respond to further questions.

Unlike most news websites, La Tilde carries no bylines, masthead, or mention of actual staff of any kind. Although the site claims it employs “dozens of freelance reporters and content creators,” at least some of the site appears to have been generated by a large language model. Running articles through Pangram, an AI-text detection service, produced multiple hits for both English and Spanish writing either partially or entirely written by machines (though such tools are known to deliver false positives).

Emerson Brooking, a fellow with the Atlantic Council’s Digital Forensic Research Lab and former Pentagon cyber-policy adviser, told The Intercept he was struck by site’s shoddiness, describing it as “AI all the way down.”

Despite the low quality of AI-generated articles, this approach could help the Pentagon spin up propaganda efforts faster than in the past. “If you can generate new content and even news fronts at the flip of a switch, your influence operations can shift target and focus much more quickly,” Brooking said. “That seems to be the thinking behind recent AI-powered Russian and Chinese networks, for instance.”

An analysis of subdomains hosted on LaTilde.co reveals the site plans to launch bespoke versions for readers in Ecuador, El Salvador, Guyana, Honduras, Jamaica, Panama, and Peru.

Some pro-U.S. content is clearly tailored to these national audiences. An article filed to the site’s “In Good Hands” section highlights the benefits of U.S.–Panamanian joint jungle warfare training exercises, regaling readers with how “temperatures and heart rates climb at the Cristóbal Colón Naval Air Base as Panamanian security forces push forward through the ‘Green Mile,’ the demanding final test of the Combined Jungle Operations Course.” Such joint initiatives are, according to La Tilde, a bulwark against China’s efforts to engage in similar joint exercises in Latin America. Rather than engage with “Beijing’s predatory practices,” the article suggests countries should follow Panama’s lead and “seek training opportunities closer to home or with longstanding partners such as the United States.”

Related

Pentagon Reveals Attacks in Latin America Are Just the Beginning

The article makes no mention of the controversy surrounding PANAMAX, a joint military exercise between SOUTHCOM and the Panamanian forces that has sparked increased protest on the grounds it violates national sovereignty. Permanent U.S. military installations in Panama were shuttered in 1999 as part of a 1977 treaty between the two countries; Panamanian opposition parties decried the reestablishment of an American military presence under the guise of joint exercises as a “camouflaged invasion.” Participants in the 2025 PANAMAX exercise La Tilde is pushing include the Western Hemisphere Institute for Security Cooperation, previously known as the School of the Americas, a Pentagon training institute whose graduates included thousands of Latin American death squad gunmen and dictator Manuel Noriega.

The importance of military and intelligence-sharing compacts with the U.S. is a recurring theme. “Far from weakening sovereignty, this kind of cooperation can strengthen it,” one article says.

Other stories from La Tilde argue the American side of Latin American controversies, similarly downplaying issues of national sovereignty. One piece describes how the U.S. abduction of Maduro “has reawakened a long-contained hope among millions of Venezuelans inside and outside the country.” Another alleges Ecuador is a nexus of the international cocaine trade, echoing claims the Trump administration has used to expand Operation Southern Spear, SOUTHCOM’s Caribbean airstrike campaign that has killed more than 200 civilians to date.

It’s unclear who exactly is operating the site on a day-to-day basis. A similar network of military propaganda pages, descendants of an Obama-era information warfare program called the Trans-Regional Web Initiative, appears to be administered by military contractor General Dynamics Information Technology. Renée DiResta, who co-authored a 2022 report on online propaganda efforts backed by U.S. Central Command, told The Intercept that the TRWI successor websites share a common Google Ads identifier code owned by General Dynamics, according to a recent comprehensive analysis of the network she conducted. La Tilde also runs a legal disclosure with identical language as those sites.

General Dynamics did not respond to multiple requests for comment about La Tilde.

Halcyon Group International, another information warfare contractor that operates Diálogo Américas, a similar pseudo-news site backed by the Pentagon, told The Intercept it was not involved with La Tilde.

Design of the La Tilde website was subcontracted to Antpack, a Colombian digital marketing firm. Multiple files hosted on the site created by the AI image-generation service Midjourney contain the word “Antpack” in their name. The Intercept signed up for a user account on La Tilde, part of planned functionality that will let readers comment and save articles for later. Once registered, The Intercept was able to view comments left on a non-public version of the site used by its developers, who posted under names corresponding to LinkedIn profiles of Antpack employees. Antpack did not respond to multiple requests for comment.

Related

Pentagon Document: U.S. Wants to “Suppress Dissenting Arguments” Using AI Propaganda

U.S. Special Operations has a long record of leading the American internet propaganda efforts, ranging from high-tech efforts to less-sophisticated projects like phony online newsrooms. SOCOM has since 2018 operated the Joint Military Information Support Operations Web Operations Center, which coordinates information warfare and online psychological operations.

The Intercept reported in 2023 that SOCOM was working on acquiring state-of-the-art “deepfake” video fabrication technologies to “generate messages and influence operations via non-traditional channels,” according to procurement documents. La Tilde appears to be using low-effort AI tools rather than anything cutting-edge. Art accompanying its stories often includes portion of the prompt used to quickly generate the image in the file name, and shows mixed results, such as a rendering of the White House portico missing several of its columns or a diploma with garbled text. Photographs illustrating pro-SOUTHCOM messaging, however, are drawn from the Defense Visual Information Distribution Service, an official Pentagon media library.

“The intent is probably to fill these sites with generic material, build an audience base, and then slip in more pieces of explicit propaganda, like that rather fulsome recounting of the U.S. attack on Venezuela,” Brooking said. “This is how you build these sorts of networks. But the content is lazy, the AI is bad, and the required disclosures make the whole thing a farce.”

The post The Pentagon Is Running an AI Propaganda Mill Targeting Latin America appeared first on The Intercept.

Musk’s break with OpenAI, which he co-founded in 2015, is in a sense a lawsuit about safety. He contends that Altman betrayed the company’s original nonprofit mission of safely and responsibly pursuing artificial intelligence for the public benefit by converting it into the revenue-maximizing behemoth it has become. According to Musk, the stakes of this are existential for the human race: “It could kill us all,” he testified on Tuesday. “We don’t want to have a ‘Terminator’ outcome.”

The AI safety community frequently invokes these dystopian scenarios to both warn the public about the technology’s risks and implicitly boast of its great power. While such a science-fiction future may lay ahead, these warnings overlook the deadly present. Artificial intelligence is already targeting humans with the blessing of Musk and his rivals.

Musk and others who caution about an uprising of sentient killer machines are anticipating the emergence of “artificial general intelligence,” an ill-defined form of superior machine reasoning that may never come to pass. But their fear that AI could kill us all is less hypothetical for those living in places targeted by the Trump administration’s global wars. In Iran, for instance, Anthropic’s Claude AI model “suggested hundreds of targets, issued precise location coordinates, and prioritized those targets according to importance,” according to the Washington Post.

“The risks of integrating frontier AI into the nation’s most lethal capabilities are already existential, both for civilians swept up in the violence and destruction of AI-enabled wars, and rank-and-file troops that have to live with the consequences of potentially unsafe weapons they can’t control,” Amoh Toh, senior counsel at Brennan Center’s Liberty and National Security Program, told The Intercept. “Existing AI models are already pushing policymakers and militaries toward nuclear escalation — there’s a real danger of Skynet-like outcomes even without a Skynet-style takeover.”

Related

OpenAI on Surveillance and Autonomous Killings: You’re Going to Have to Trust Us

Silicon Valley has widely embraced AI military contracts despite its worries over lethal AI. Amazon, OpenAI, Musk’s xAI, and Microsoft all earn money from selling large language model services to the Pentagon. Even Anthropic, accused of “betrayal” by War Secretary Pete Hegseth and declared a national supply chain risk for mounting the smallest of opposition to the Pentagon’s terms, is still keen to participate in the national kill chain. “Anthropic has much more in common with the Department of War than we have differences,” CEO Dario Amodei wrote in a blog post a week after the United States bombed an elementary school in Iran, killing more than 100 children.

Google offers a telling illustration of the industry’s increasing coziness with selling AI to the military. Following a 2018 employee revolt over Project Maven, a contract to help target Pentagon airstrikes, CEO Sundar Pichai pledged his company would swear off the business of killing. He wrote in a company blog post that Google would not pursue deals that could cause harm, including applications whose “principal purpose or implementation is to cause or directly facilitate injury to people.” He added: “These are not theoretical concepts, they are concrete standards that will actively govern our research and product development and will impact our business decisions.”

After watching AI help wage a war that has already killed over 1,700 Iranian civilians, Google this week sent a clear message: We want in. In a deal that makes explicit the extent to which company leadership has abandoned its AI principles, Google agreed to provide AI services to the Pentagon that allow for “classified workloads,” sensitive military work that encompasses tasks like intelligence analysis and targeting airstrikes, The Information reported.

According to the tech news outlet, the deal allows the U.S. military to use Google’s AI models for “any lawful government purpose” — a carveout that could allow any uses the administration deems legal. Take, for example, the Trump administration’s Operation Southern Spear, the ongoing aerial assassination program against civilian boats accused of drug trafficking that has killed more than 180 people to date. The campaign has been widely condemned as illegal under both international and U.S. law, but the administration has deemed its own actions legal through a Department of Justice memo that remains secret. On Friday, the Pentagon announced additional “lawful operational use” deals with Nvidia, Microsoft, and Amazon as well.

The Google contract reportedly includes a toothless and unenforceable provision gesturing at concerns over autonomous and spying. “We remain committed to the private and public sector consensus that AI should not be used for domestic mass surveillance or autonomous weaponry without appropriate human oversight,” the clause reportedly states.

“When I worked at Google, they would spend a lot of time punting into the future, promising a future that would never come,” said William Fitzgerald, a former Google employee who helped organize the 2018 worker-led campaign against the Maven contract. “‘Don’t regulate us or it’ll kill innovation.’ The talking point is the same today. The reality of Google’s work with the military is it’s part of a tech-military ecosystem that’s killing people today.”

Google spokesperson Kate Dreyer did not respond to questions about the contract’s language, instead touting how the company’s military work applies “to areas like logistics, cybersecurity, diplomatic translation, fleet maintenance, and the defense of critical infrastructure.”

There is little evidence the people in charge find this technology enticing because of its diplomatic translation prowess. In a January address to Musk’s employees at SpaceX, another Pentagon contractor, Hegseth explained how “an embrace of AI” would make the military “more lethal.”

Related

AI’s Imperial Agenda

Musk and Altman, though foes at the moment, can at least find common ground in their support of Hegseth. Musk, a longtime defense contractor, similarly wraps himself in the flag, tweeting in 2023, “I will fight for and die in America.” Altman, who once expressed skepticism toward military work, now frames OpenAI’s mission in terms of patriotic nationalism. (In 2024, The Intercept sued OpenAI in federal court over the company’s use of copyrighted articles to train its chatbot ChatGPT. The case is ongoing.)

Between Musk’s courtroom visions of the apocalypse and Google’s plunge into classified workloads, the week’s news illustrates the disjointed state of AI industry ethics, where executives say they’re terrified of the technology killing by accident, while wholly supportive of using it to kill on purpose.

Though AI executives clearly find this a virtuous revenue stream, some of the people who actually built the technology do not. Andreas Kirsch, a research scientist at Google’s pioneering DeepMind laboratory that produced much of the work on which xAI and Anthropic rely, responded to this week’s news with dismay: “I’m speechless at Google signing a deal to use our AI models for classified tasks. Frankly, it is shameful,” he wrote on X. Alex Turner, a DeepMind colleague of Kirsch’s, described the contract in a single word: “Shameful.”

The post Musk Warns of Killer AI — While He and the Rest of Silicon Valley Cash In on AI That Kills appeared first on The Intercept.

Since 2018, the Internal Revenue Service’s Criminal Investigation division has used Palantir’s Lead and Case Analytics platform to aggregate and analyze a sprawling list of sensitive federal databases and data sets.

Public records detailing Palantir’s IRS contract, obtained by the nonprofit watchdog group American Oversight and shared exclusively with The Intercept, reveal the immense volume of data plugged into the military contractor’s software. The LCA uses both Palantir’s Gotham and Foundry applications to facilitate “analysis of massive-scale data to find the needle in the hay stack,” the contract paperwork says.

Documents indicate the IRS has paid Palantir over $130 million for these services to date.

Palantir’s LCA is ostensibly directed toward cracking down on fraud, money laundering, and other financial crimes. According to a 2024 agency privacy impact assessment, IRS “Special agents and investigative analysts … utilize the platform to find, analyze, and visualize connections between disparate sets of data to generate leads, identify schemes, uncover tax fraud, and conduct money laundering and forfeiture investigative activities.”

Related

Trump Wants to Put You in a Massive, Secret Government Database

The IRS use of the software, launched under Trump’s first term and expanded under Biden, is now in the hands of an IRS Criminal Investigations office that has drastically scaled back its pursuit of tax cheats and pivoted, under Trump’s direction, toward investigating “left-leaning groups,” the Wall Street Journal reported in October.

“The real concern is the consolidation of vast amounts of sensitive personal data into a single system with minimal transparency — especially one built and operated by a contractor like Palantir, whose business model is premised on integrating data and expanding surveillance capabilities,” American Oversight director Chioma Chukwu said in a statement to The Intercept. “Its platforms have been used in deeply troubling contexts, from immigration enforcement to predictive policing, with persistent concerns about overreach, bias, and weak oversight.”

Palantir did not respond to a request for comment, nor did the IRS.

The contract documents reviewed by The Intercept reveal that these “disparate sets of data” are vast. Palantir’s LCA allows the IRS to quickly search and visualize “connections from millions of records with thousands of links” between databases maintained by the IRS and other federal agencies. According to the contract documents, this data includes individual tax form and tax returns as well as Affordable Care Act data, bank statements, and transactions, and “all available” data compiled by the Treasury Department’s Financial Crimes Enforcement Network.

Its view apparently extends to cryptocurrencies including bitcoin, Ethereum, Litecoin, and Ripple. “The application would sit on top of a singular repository of identified wallets from seized servers utilizing dark web data obtained from exchangers such as Coinbase,” the documents note.

The program places an emphasis on mapping social relationships between the targets of an investigation. That includes analyzing a “network of people and the relationships and communications between them,” such as “calls, texts, [and] emails events.” The use of “IP address analysis” within LCA allows the IRS to “Identify suspects more easily” and “Establish (new) relationships among actors.”

These investigative functions are continuously updated, the materials say, through ongoing close work between Palantir engineers and IRS personnel.

Related

Palantir Will No Longer Profit Off of New Yorkers’ Health Data

The intermingling of sensitive data on millions of Americans comes at a time of increased global skepticism and opposition toward Palantir, which, despite its military-intelligence origins, has a thriving business with civilian agencies like the IRS. The use of Palantir software at the U.K.’s National Health Service, for example, has created an ongoing political controversy across Britain, while a similar contract with the New York City public hospital network was recently canceled following public protest.

The contract is also active at a time when IRS Criminal Investigations has been coopted to aid in the broader Trump administration’s aggressive agenda. In July, ProPublica reported that the agency was working with U.S. Immigration and Customs Enforcement to provide “on demand” data to accelerate deportations. Last year, the New York Times reported that Palantir, founded by Trump ally Peter Thiel, was central to an administration effort to increase data-sharing across federal agencies.

The company’s right-wing politics and eagerness to facilitate U.S. and Israeli military aggression abroad, NSA global surveillance, and ICE deportations has also made many wary of its access to incredibly sensitive personal data. A recent post on the company’s Palantir’s X account summarizing a book by CEO Alex Karp triggered an immediate backlash from those unnerved by the manifesto’s fascistic bent. The bullet points extolled the virtue of arms manufacturing, argued the Axis powers were unfairly punished after World War II, called for a reinstatement of the draft, condemned cultural pluralism, and claimed that wealthy elites are unfairly persecuted.

“When the government can map relationships, track behavior, and generate investigative leads across data sets at this scale, the question isn’t just what it can do — it’s who it will be used against,” Chukwu said. “Entrusting that infrastructure to a company known for opaque, security-state deployments only heightens those risks.”

The post Palantir Is Helping Trump’s IRS Conduct “Massive-Scale” Data Mining appeared first on The Intercept.

Last year, the LAPD launched its “Drone as First Responder” program with a clearly articulated goal: to protect and even save lives. The pilot program authorized the rapid deployment of drones to the scenes of certain emergency calls before human officers even arrive. After receiving a 911 call, authorities can dispatch a drone to get a better picture of what’s happening from the sky, potentially reducing the number of officers dispatched. This means police resources could, theoretically, be more efficiently deployed to other emergencies around the city.

“This innovative program not only aims to enhance transparency in Department operations but also prioritizes the protection of individual privacy,” the LAPD explained in a webpage about the program. “By deploying drones as an invaluable resource for patrol officers, the DFR Pilot Program provides a cutting-edge tool that can respond swiftly to emergencies, ensuring a safer environment for all.”

The LAPD turned to Skydio, a California-based drone startup that previously marketed its aircraft to consumers but has pivoted to supplying militarized, weapons-compatible hardware for the U.S. Army, Israeli Defense Forces, and other governments.

The LAPD insists the DFR program presents no threat to personal privacy or civil liberties. “Unless you are in the commission of a crime or under criminal investigation for the commission of a crime,” assures the website, “the officers utilizing the drone are not interested in recording you.”

Related

How to Track Kash Patel’s Jet

But according to flight data shared publicly by the LAPD and Skydio, the city has used DFR not only to respond to emergencies, but also to monitor multiple protests across Los Angeles. Software engineer and flight data researcher John Wiseman has tracked DFR aircraft to at least two protests in Los Angeles this year, he told The Intercept, raising questions as to whether the city is operating an aerial surveillance program against nonviolent, constitutionally protected activity.

Flight records show DFR drones were launched at least 31 times to surveil the January 31 “ICE Out” protest in downtown Los Angeles, which saw thousands peacefully march against the administration’s deportations raids and street violence in Minneapolis. The Los Angeles Times said the “mostly peaceful protest took a turn as day turned to night in downtown Los Angeles and the crowd refused to disperse,” whereupon police began firing tear gas at remaining demonstrators.

At the March 28 “No Kings” protest against the Trump administration, city data shows the LAPD again launched drones 32 times over the area where the demonstration took place. A heat map visualization created by Wiseman based on the city data shows the drones lingered for extended periods over the Metropolitan Detention Center and the intersection of North Central Avenue and East Temple Street in Los Angeles’s Little Tokyo neighborhood.

Following the protest, the city’s local ABC News affiliate reported the event “drew tens of thousands who listened to speakers before marching peacefully through downtown streets.” The LAPD later arrested 75 individuals, 74 of whom were taken in simply for not dispersing when ordered by police.

The DFR flight data shows the drones began orbiting the protest at 2 p.m., hours before the order to disperse was issued at 5:30 p.m., and continued flying until 9 p.m. that evening. Nine drone flights began before the dispersal order.

In response to questions about the protest surveillance, LAPD Lt. Matthew Jacobs told The Intercept, “We do not document or record unless there is a crime occurring.”

“When it comes to a protest or demonstration, we’re responding [with drones] at the request of the Incident Commander,” Jacobs said. “We’re looking for specific people, we’re not taping First Amendment activity.”

Related

LAPD Won’t Do Immigration Enforcement — But Will Shoot You With Rubber Bullets for Protesting ICE

Jacobs added that “99 percent of the time” drones are sent to a protest “because the commander reports a crime in progress,” and claimed a “wide variety of crimes” are committed at protests, from vandalism to rocks thrown at officers. Jacobs added at times the department simply “wants to see how big a crowd is.”

When asked why drones were surveilling the No Kings protest hours before the dispersal order, Jacobs said that the LAPD “cannot provide deeper insight into specifics of a single flight.”

When not recording, Jacobs said DFR cameras are monitored by both their pilots and LAPD personnel on the ground, who have access to the live feeds. Any recorded footage is stored on an indefinite basis.

The police department did not answer a detailed list of follow-up questions, including how much protest-related data it has captured via drone surveillance to date or who monitors drone feeds over protests.

The LAPD’s fleet of Skydio X10 drones monitor the ground using with a sophisticated suite of sensors the company says are capable of detecting the presence of person from a distance of more than 8,000 feet and identifying an individual more than 2,500 feet away. The company also touts the drone’s ability to read license plates from a distance of 800 feet. Last year, Skydio CEO Adam Bry demonstrated how two police officers using the company’s DFR Command software could operate eight drones at once between them, tracking license plates and automatically following people of interest.

Update: April 20, 2026, 4:08 p.m. ET
This article was updated to include new comment from the Los Angeles Police Department.

The post LAPD Deployed Drones to Spy on No Kings Protest appeared first on The Intercept.

These overlooked stories share the same ideological slant and editorial voice: that of the White House. Al-Fassel’s YouTube account, for instance, has racked up millions of views on Arabic-language videos praising the Trump administration’s Gaza policy and exhorting Hamas to cease “taking orders from the Iranian regime” and release Israeli prisoners. On Pishtaz News, a poll on the homepage recently asked: “[H]ow would you describe your belief about the Supreme Leader’s current health status and whereabouts?” Possible answers range from “In good health but hiding” to “Disfigured” or “Dead.” The excellence of Saudi and Emirati leadership, both close military partners of the U.S., is a recurring theme.

There’s a reason this coverage echoes American foreign policy talking points. Al-Fassel and Pishtaz News are, in fact, part of network of websites and social media accounts purporting to be legitimate Middle Eastern news outlets that are in fact propaganda mills funded by the United States government, The Intercept has found.

Disclosed only at the bottom of both sites behind an “About” link that is easily missed by casual readers, the outlets note that they are “a product of an international media organization publicly funded from the budget of the United States Government.” The government affiliation remains undisclosed on social media platforms including Instagram, despite a platform policy requiring the labeling of state-backed media outlet to prevent the unwitting consumption of government propaganda.

The sites’ recent fixation on crushing Iran is unlikely to be a coincidence: Both publications share numerous connections with a portfolio of fake newsrooms that originated as a military psychological operations campaign against foreign internet users.

Al-Fassel and Pishtaz News did not respond to requests for comment, nor did CENTCOM or the Department of Defense.

In 2008, U.S. Special Operations Command put out a call for contractors to help operate what it called the Trans-Regional Web Initiative, a project that would provide “rapid, on-order global dissemination of web-based influence products and tools in support of strategic and long-term U.S. Government goals and objectives.” In other words, state propaganda pushed by Pentagon.

Masquerading as independent online newsrooms, the TRWI sites hired “indigenous content stringers” to produce articles “which Combatant Commands (COCOMs) can use as necessary in support of the Global War on Terror.” The contract, awarded to General Dynamics Information Technology, spawned 10 websites that funneled U.S. foreign policy talking points to audiences across the Middle East and South Asia, running everything from banal essays about inter-faith coexistence to, as reported by Foreign Policy in 2011, articles intended to “whitewash the image of Central Asian dictatorships.” By 2014, the sites were deemed a failure by Congress and de-funded.

Eight years later, a team of researchers published an unusual report. Following the 2016 election, the bulk of the Western media’s interest in online propagandizing had focused on influence campaigns attributed to Russia, China, and other American geopolitical rivals. But the 2022 report from the Stanford Internet Observatory and Graphika, a commercial internet analysis firm and Pentagon information warfare contractor, uncovered a network of phony “pro-Western” Twitter and Facebook accounts that pushed articles from pseudo-news websites. The report stopped short of formally attributing the campaign to the U.S., but noted that both Meta and Twitter had done so. The researchers concluded that the accounts in question attempted the coordinated spread of articles from a network of sham news websites established by U.S. Special Operations Command.

Related

Twitter Aided the Pentagon in Its Covert Online Propaganda Campaign

The report found that just a few years after TRWI’s ostensible death, many of the sites had simply rebranded, now carrying hard-to-find disclosures mentioning they were run by U.S. Central Command. Following Stanford and Graphika’s findings, some of the sites shut down; others continued. Subsequent reporting by the Washington Post found that the embarrassing revelations spurred the Pentagon to conduct “a sweeping audit of how it conducts clandestine information warfare.”

A review of the Internet Archive shows that in the aftermath of the Stanford report, TRWI sites that remained in operation changed their disclosure language. Rather than citing CENTCOM sponsorship, these sites shifted to state that they are “publicly funded from the budget of the United States Government.” The disclosure language used by the remaining network of CENTCOM propaganda sites is a word-for-word copy of the phrasing The Intercept found tucked away on the About pages of Pishtaz News and Al-Fassel.

That’s not the only evidence suggesting a link to this network of military propaganda sites.

Since they began publishing in 2023, Al-Fassel and Pishtaz News have regularly quoted or summarized CENTCOM press releases touting regional operations and battlefield successes, as did the outlets mentioned in the Stanford/Graphika report. The reliance on combatant command press releases in particular is an editorial strategy that dates back to the original SOCOM-run TRWI network.

On X, Pishtaz News follows only three other users; two are the official CENTCOM accounts for Farsi and Arabic audiences. The Pishtaz News Instagram account, which carries no disclosure of the account’s governmental nature, follows only one other user: “US CENTCOM FARSI.”

Intentionally or otherwise, Al-Fassel’s posts to X are often geotagged as having been sent from Lutz, Florida, a stone’s throw from the headquarters of CENTCOM and SOCOM in Tampa, as well as myriad military contractors that service both.

Both sites also share common design elements with the TRWI-associated publications that suggest they were created or operated by the same contractor: All posts conclude with a poll asking “Do you like this article?” using the same thumbs-up and thumbs-down icons. URLs are structured identically for Al-Fassel, Pishtaz News, and Salaam Times — an Afghanistan-focused site launched under the TRWI that continues today under a different name — suggesting they were coded using the same tools. The three sites use an identical 404 error graphic to alert users when they’ve clicked on a broken link, as well.

The web design of Al-Fassel and Pishtaz News — including page layout, URL structure, 404 error graphic, and much of the legal verbiage in the About sections — closely mirrors that of CENTCOMcitadel.com, a publication with similar content that carries an overt disclosure of Pentagon sponsorship at the bottom of its homepage.

“These sites are similar in style to the overt messaging efforts we saw from the Department of Defense previously,” Renée DiResta, a former Stanford researcher and co-author of the 2022 report, told The Intercept. “We previously saw this pattern of clearer U.S. affiliation language in the About page of the domain, then minimal to no acknowledgement on the social media profiles.”

There are other subtle nods to the sites’ true purpose: URLs for the English language versions of each site are denoted “en_GB,” for Great Britain. In a comprehensive 2015 analysis of the TRWI network, University of Bath doctoral student Roy Revie observed that the network of American military propaganda sites explicitly marked their English versions as British because “SOCOM seeks to avoid any suggestion its sites are aimed at US audiences.”

In the parlance of information warfare, these propaganda shops are considered “overt” rather than “covert,” because their state ownership is technically disclosed. But in his 2015 paper, Revie argued that these psyop sites still engage in deception. They use online journalism as a form of camouflage, he wrote, because most readers won’t seek out a publication’s About page to learn about its funding. The design of these sites “allows the DOD to credibly claim full transparency and maintain legitimacy, putting the onus onto the user to inform themselves about the source,” Revie wrote.

The output of both sites consistently lionizes the U.S. and Israel, along with America’s Gulf allies. They regularly demean the Iranian state, presenting a wholly lopsided and misleading account in a time of war. “The US says it does not seek open conflict with Tehran,” reads a March 2 article in Al-Fassel. Both sites have repeatedly cited reporting by Iran International — a Saudi-funded, pro-Israel, Iranian monarchist publication with a long record of journalistic misrepresentation. A March 31 Pishtaz News article, for instance, based on an entirely anonymously sourced Iran International post, alleged that Iranian security forces gang-raped nurses in Tehran.

Recent coverage depicts Iran as up against the ropes. A March 22 article in Pishtaz News exclaimed, “The Islamic Republic’s regular army, known as the Artesh, is increasingly described by informed observers as a force under severe strain and institutional neglect.” Another anonymously authored piece from March 25, headlined “Artesh would be better off without its main rival,” seems intended to stoke tensions between Iran’s regular army and its Islamic Revolutionary Guard Corps. “Without the Islamic Revolutionary Guard Corps (IRGC), resources could flow directly to the regular army, known as the Artesh, enabling meaningful modernization,” the story claimed, a talking point ripped straight from the mouths of right-wing Iran hawks in the U.S. In a March 18 Fox News segment, for example, retired Gen. Jack Keane suggested that an Artesh–IRGC rivalry could be exploited to accomplish regime change.

It’s unclear who exactly writes what appears on these sites. Most articles run without any byline, while other stories are published under names that are difficult to find any mention of anywhere else on the internet. Some of the personnel may not be real at all. A January Al-Fassel YouTube overview of recent regional headlines was narrated by an Arabic-speaking man in a sharp blue blazer. Experts told The Intercept the newscaster was likely a product of generative AI and not genuine footage. “The strongest indicator is an almost complete absence of eye blinks,” Georgetown University professor and deepfake researcher Sejin Paik told The Intercept. Zuzanna Wojciak, a synthetic media researcher with the human rights organization Witness, reached the same conclusion, citing strange anomalies with his skin, hands, and teeth.

Related

Pentagon Document: U.S. Wants to “Suppress Dissenting Arguments” Using AI Propaganda

Some articles deeply misstate or misrepresent the facts. An April 15 Al-Fassel article about Iran’s “war crime threats” against the American University of Beirut omitted the fact that these threats came in response to repeated U.S.–Israel airstrikes against Iranian schools. The day after an Al-Fassel article described the Houthis as “crippled” and “largely disintegrated,” capable of offering only “verbal support” for Iran, the Yemeni militant group launched cruise missiles at Israel.

The outlets also illustrate the extent of deceptive messaging radiating from the Pentagon and White House: A March 5 post to the Pishtaz News Instagram account boasted, “The Iranian regime’s ability to strike US forces and regional partners is rapidly eroding, while US combat power continues to grow.” Four weeks later, Iran was continuing to lob missiles at U.S. bases as well as its regional partners, and succeeded in downing an American F-15 and A-10 Warthog. An April 4 Al-Fassel Instagram post claimed, citing Secretary of State Marco Rubio, that “Iran is not satisfied with a peaceful nuclear program, but seeking to enhance its military capabilities,” even though a 2025 assessment from the Office of the Director of National Intelligence concluded the opposite.

Other articles dispense with masquerading as journalism, reading more as warnings straight from Washington: “United States is fully prepared to protect its forces in Middle East,” read a June 2025 headline on Pishtaz News. “With advanced technological capabilities and highly-trained personnel, the United States maintains one of the world’s most capable military forces, continuously adapting to evolving security challenges to maintain order and stability.” A March 27 Pishtaz News tweet was more straightforward. “You will be systematically annihilated,” it threatens in Farsi. “Your commanders are hiding in bunkers. They have sent their families and wealth abroad—why are you still fighting for them?”

Some articles purport to include comments from genuine expert sources. In at least one case, this happened without the knowledge of the source. A July 2025 article in Al-Fassel predicted that a future closure of the Strait of Hormuz “would harm China and Russia more than other nations.” The article quoted Umud Shokri, an energy analyst affiliated with George Mason University, the State Department, and the Middle East Institute. “I would like to clarify that I was not aware of any affiliation between alfasselnews.com and the U.S. government,” Shokri told The Intercept. “I also did not have any direct interview with the platform, nor was I contacted by them directly. To the best of my knowledge, any quotation attributed to me appears to have been drawn from prior public commentary or other media appearances.”

Prior to the war on Iran, a top priority on both sites was marketing the U.S.–Israeli plans for the future of Gaza. The message is essentially a distillation of the U.S.–Israel–Gulf State consensus: That all Palestinian suffering is brought on by Hamas rather than the past three years of Israeli bombardment, and that the Trump-sponsored “Board of Peace” augurs an unprecedented era of prosperity for Palestinians.

Related

Plans Call for “New Rafah” Built in Israel’s Image — Without Palestinians

“The incoming Board of Peace,” a December 2025 Al-Fassel piece claimed, “is expected to foster conditions for democratic representation and meaningful civic participation.” A December 12 Al-Fassel YouTube video similarly blamed Hamas and Iran, rather than Israel, for the blockade of humanitarian aid into Gaza, followed by an AI-generated image of a science fiction city overlaid with Arabic captions promising billions in foreign investment and economic revitalization for Gaza. The video currently has nearly 1.7 million views.

Other items around Gaza further invert reality. Since October 2025, Gaza has been bifurcated by the so-called “Yellow Line,” an arbitrary boundary behind which Israeli forces nominally withdrew last year. Palestinians on the Israeli side of the line face harsh occupying military governance, while those on the other side risk being killed.

Despite claims by Al-Fassel’s video team that Trump’s Gaza policy will herald the ability for countless Palestinians to return home, Israeli forces routinely fire at civilians approaching this buffer zone.

“Incidents of gunfire, shelling, and limited incursions have continued near the ‘Yellow Line,’ the separation zone near the border with Israel, keeping any return highly dangerous,” according to a United Nations video report. “With the amount of available space shrinking, thousands of families have been forced to return to the edges of their destroyed neighborhoods near the ‘Yellow Line,’ despite what residents say is the continued risk of injury or death from intermittent fire.”

Not so, says Al-Fassel: “The Yellow Line is more than a boundary; it is a lifeline designed to keep Gaza’s families safe and informed during the ceasefire,” claimed a November article. “The Yellow Line is not a symbol of division — it is a lifeline.”

Following the 2016 election and the panic surrounding Russian covert propaganda efforts, major American social media platforms began adding labels to the accounts of government-controlled media properties. Videos from Al Jazeera English’s YouTube account, for instance, come with a disclaimer that “Al Jazeera is funded in whole or in part by the Qatari government.” Although X abandoned this policy in 2023, it is still nominally on the books for both Meta, which owns Facebook and Instagram, and YouTube.

Related

A Journalist Reported From Palestine. YouTube Deleted His Account Claiming He’s an Iranian Agent.

There is no disclosure, however, in the Instagram posts or accounts of Al-Fassel or Pishtaz News. YouTube videos from both accounts do not include a disclaimer about U.S. funding; however, a brief disclosure can be found on their main account pages, tucked into an About section that must be expanded to be read.

Neither site appears to have a particularly large audience on social media. Both have paltry followings on X — about 2,400 for Al-Fassel, and only 132 following Pishtaz News — with many appearing to be spam-based accounts with names followed by a long string of numbers that engage in posting behavior common to spam networks. Al-Fassel has found modest engagement on Instagram, where it has over 7,700 followers. Though Pishtaz News has only 475 followers on Instagram, its posts sometimes break through; a March 18 post of CENTCOM footage from the deck of an aircraft carrier, for example, racked up more than 1,100 likes.

At times, the content published by the propaganda sites may have reached American audiences. A March 27 Al-Fassel story alleging the total collapse of the Iranian-led “Axis of Resistance” was shared that same day to FreeRepublic, the conservative American message board, by user MeanWestTexan. Federal law forbids Pentagon propaganda aimed at Americans, though a similar prohibition aimed at the State Department was overturned in 2013.

Sometimes their stories reach other Western readers. An Al-Fassel article on the Houthis made its way into the citations of a 2024 article in the academic journal Survival: Global Politics and Strategy by University of Ottawa professor Thomas Juneau. (Juneau did not respond to a request for comment.) A submission to the U.N.’s Committee on Enforced Disappearances from Justice for All International, a Swiss-based nonprofit, similarly cited an Al-Fassel post on the IRGC, while an annual report by the state-operated Swedish Defence Research Agency relied in part on an Al-Fassel article on ISIS. The Intercept reviewed multiple entries on Grokipedia, X’s Wikipedia clone, citing Al-Fassel articles as well.

Emerson Brooking, a fellow with the Atlantic Council’s Digital Forensic Research Lab and former Pentagon cyber policy adviser, believes CENTCOM is most likely behind the sites and considers their overall reach lackluster. When it comes to online propaganda, he said, the U.S. “could learn some lessons from Iran.” Iranian propaganda efforts — mostly quickly produced AI slop — have captured the attention of the internet in a way that the U.S. ersatz newsrooms have not.

But the sites’ limited reach is unlikely to bring them to a halt anytime soon. Even as the Trump administration has gutted Voice of America and other long-standing tools of U.S. soft power, these sites have continued publishing. If their similarities to the long-running American military psyops are more than coincidental, that says more about a culture of inertia at the Pentagon than its success in winning hearts and minds. Brooking told The Intercept that because operating blogs amounts to a “rounding error” within the broader defense budget, such projects can continue with little scrutiny.

A seldom-read network of propaganda sites might seem to have little purpose. But it’s the kind of thing authorities can gesture toward, Brooking said, when pressed about their efforts to combat Iran in the “information space.” “Successive SOCOM or CENTCOM or other senior leaders could point to the fact that they’re maintaining this network of websites,” he said.

The post These Middle Eastern News Sites Are Actually U.S. Government Propaganda Operations appeared first on The Intercept.

This spring, Meta quietly revised its Community Standards policy, an internal company document dictating what its billions of global users can and cannot say online. The latest tweaks can be found in a chapter on “Violence and Incitement,” where a subsection titled “Other Violence” spells out, among other rules, the company’s bans on ads for assassins. It’s in this subsection where Meta last month published a revision to include new limitations for users who mention antifascism.

Policy documents reviewed by The Intercept show the company now treats any “Content that includes the word ‘antifa’ as a potential rules violation if that word appears along with what Meta deems a “content-level threat signal” — meaning a statement that the company believes implies violence.

In some cases, the content that Meta considers a threat signal is commonsensical. If, for instance, a user mentions bringing a weapon to an event, the company flags it as a threat signal. But in other cases, Meta’s process for identifying threat signals is more vague. Under the new rules, Meta might trigger a threat signal when a user posts a “visual depiction of a weapon,” a “reference to arson, theft, or vandalism,” or “military language,” if accompanied by the word “antifa.”

If “antifa” is mentioned in the context of “references to historical or recent incidents of violence” — a category so sprawling that it includes “historic wars” and “battles” —  that post will also be penalized. Should Meta apply this rule as written, the company could, for instance, restrict posts comparing the antifascist nature of World War II to the contemporary antifa movement.

Potential penalties for violating Community Standards range from a full account ban to comments being hidden or suppressed.

The policy change follows years of Meta and its chief executive Mark Zuckerberg’s pivot of political convenience toward President Donald Trump and his base. Following Trump’s second electoral victory, Meta quickly changed its speech rules to allow for anti-transgender slurs and dehumanization of immigrants, The Intercept previously reported, aligning the company with longtime MAGA culture war grievances.

Related

A Redditor Criticized ICE. Trump Is Trying to Unmask Them by Dragging the Company to a Secret Grand Jury.

Asked about the new restrictions on the word “antifa,” Meta spokesperson Erica Sackin pointed to a March transparency report that noted the company would “remove QAnon and Antifa content when combined with content-level threat signals.” The report does not explain what those signals are. Meta did not respond when asked if the company had discussed its antifa speech rules with the Trump administration.

Meta largely outsources the enforcement of its Community Standards rules to low-paid contractors whose interpretation and application of the policies can vary. The company’s automated, algorithmic content moderation systems are also famously glitchy. This combination can result in erratic censorship, particularly when political ideology is classified as violent or terroristic.

The new rules around saying “antifa” on Facebook and Instagram comes amid efforts by the White House to crack down on left-wing political organizing under the guise of national security. Though antifa is a contraction of the word antifascism and not an actual group, Trump last September signed an executive order designating the leaderless decentralized movement as a domestic terrorist organization. A subsequent executive memorandum, NSPM-7, again singled out “antifa” ideology as a cause of “domestic terrorism and organized political violence.”

Prior reporting by The Intercept has shown Meta historically hews closely to federal terrorism labels. Meta in 2020 announced it would tackle the leftist bogeyman under its “Movements and Organizations Tied to Violence” policy alongside QAnon, the right-wing mass delusion that helped foment the January 6 effort to overturn the results of the presidential election by force. Though self-identified antifa adherents have taken part in acts of property damage during protests, analyses repeatedly show that left-wing violence in the United States is a relatively small and rare threat compared to right-wing extremist groups and militias.

The post Facebook and Instagram Tighten Censorship Rules for Saying “Antifa” appeared first on The Intercept.

CBP will spend up to $50 million on what it refers to as “Less Lethal Specialty Munitions,” a euphemism for weapons intended to merely hurt or disable a target rather than killing them. The agency is looking for a vendor who can supply vast quantities of 123 different types of munitions across 10 different categories, the contracting document says.

“The sheer quantity and the myriad different weapons is the most remarkable thing to me,” Rohini Haar, an emergency physician and researcher of less lethal ordnance told The Intercept. “When there’s so many different kinds, it makes you question, tactically, what’s the goal there?”

Federal agents’ indiscriminate use of “less-lethal” chemical weapons against the nonviolent demonstrators became a hallmark of the Trump administration’s immigration crackdown. Contract documents show the Department of Homeland Security will continue to stockpile a massive arsenal of tear gases and projectile weapons. (Neither CBP nor its parent agency, the Department of Homeland Security, immediately responded to requests for comment.)

Haar questioned whether the Department of Homeland Security will be able to suitably train federal agents to use such a wide variety of weapons.

“Each of them has a different sort of technical spec or specifications,” she explained. “Some of them are handheld grenades that you have to know to throw, but not hit people’s heads. Some of them are fired from a weapon, like a launcher, and so you have to be standing farther away than you would be with a grenade.”

Two Tear Gases

The shopping list includes a litany of different ways to hit people and objects with two common types of tear gas: chlorobenzalmalononitrile, or CS, a chemical weapon previously used by the U.S. in Vietnam but now banned for military use, and oleoresin capsicum, or OC, derived from chili peppers.

CBP agents already regularly use CS and OC-based weapons in the field, including against protesters. The procurement document shows that armed federal officers will continue to wield the threat of chemical agents against the public despite ample documentation of misuse.

Related

Federal Agents Used Toxic Chemical Smoke Grenades in Portland

Some of CBP’s desired weapons are designed to spread these chemical weapons indiscriminately. Included on the wish list are quart containers of liquid CS and OC meant to be spread through thermal “foggers,” dispersal devices meant to create mists with microscopic droplets of liquid. Defense Technology, a longtime chemical weapons vendor for CBP and U.S. Customs and Immigration Enforcement, says its Golden Eagle Pepper Fogger Generator can output 100,000 cubic feet of tear gas in 26 seconds.

Both chemicals are potent chemicals that can cause health effects far beyond debilitating pain.

“Greater exposure to chemical agents,” a 2023 study by the University of Minnesota School of Public Health found, “was significantly associated with higher odds of an adverse reproductive health outcomes.”

The outcomes included “uterine cramping, early menstrual bleeding, breast tenderness and delayed menstrual bleeding.”

The procurement list includes smoke grenades in four different colors and 12 different varieties of tear-gas grenades.

The weapons will be ordered in enormous volumes. CBP projects purchasing over 242,000 munitions from the “Hand Delivered Pyrotechnic Canisters” category and over 100,000 rounds of “impact munitions” fired from grenade launcher-style tubes.

The latter category includes foam-tipped “sponge cartridge” ammunition designed to either release a tear gas-style chemical upon hitting someone or merely harm them through sheer force of impact.

Maimed, Deafened, Blinded

Fired at close enough range, so-called less lethal rounds can easily kill or maim their target.

Anti-ICE demonstrator Kaden Rummler lost sight in his left eye after he was shot in the face by a federal officer in January. After the Los Angeles Police Department fired one such round directly into the face of another protester last summer, he was injured so seriously that he required surgery and had his jaw wired shut for six weeks.

“Distraction devices,” which emit loud sounds, bright lights, or other effects to stun targets, were also on CBP’s wish list, with plans to purchase 13,000 of them. The procurement document required the weapons be capable of emitting a sound of 175 decibels, louder than a gunshot or jet engine. The National Hearing Conservation Association warns of sound of 140 decibels can case permanent damage and “death of hearing tissue” begins at 180 decibels.

“In addition to injuries caused directly by the primary blast wave, such as ear-drum rupture or lung injury, secondary and tertiary injuries can also occur as a result of these explosive devices,” says a 2023 publication by Physicians for Human Rights that was co-authored by Haar.

CBP’s inclusion of rubber-ball grenades and scattershot projectiles alarmed Scott Reynhout, a researcher who also co-authored the PHR paper. When such grenades are thrown or launched at people, they release a burst of small rubber fragments akin to shrapnel in every direction and can be configured to simultaneously release tear gas.

“The procurement of the latter weapons is worrying as these have not seen widespread use yet by CBP/ICE in protests,” said Reynhout, referring to the scattershot projectiles, which he said were akin to “rubber buckshot.”

Such weapons were used by Chilean security forces against protesters six years ago, he said, resulting in more than 400 cases of partial or full-blindness, and are also employed extensively by Iranian police and paramilitaries in their crackdowns on demonstrations.

Weapons designed to pierce building materials were also included in the wish list.

CBP plans to purchase over 12,000 “ferret rounds,” projectiles filled with powdered or liquified chemicals that punch through barriers and spread tear gas on the other side.

Haar said, “If it can go through glass, particle board, and walls, it can go through a body.”

The post DHS Launches Massive “Less Lethal” Chemical Weapons Buying Spree appeared first on The Intercept.

Related

Palantir Gets Millions of Dollars From New York City’s Public Hospitals

The Intercept reported in February that the New York City Health and Hospitals Corporation, which operates a network of public health care facilities across the city, had paid Palantir almost $4 million since 2023 for data analysis services. NYCHH says it used Palantir’s software to boost its efficiency in billing Medicaid and other public benefits, which included the automated scanning of patient health notes.

The contract prompted protests from activists and local organizers who objected to the hospital system’s use of software from a company whose technology has facilitated lethal airstrike targeting, wide-reaching surveillance of American citizens, and deportation raids by Immigration and Customs Enforcement agents.

At a March 16 meeting of the New York City Council, NYC Health + Hospitals CEO Mitchell Katz disclosed that Palantir’s contract will not be renewed come October. Katz defended the health care network’s collaboration with Palantir on the grounds that there was an “absolute firewall” between patient data and the company’s government customers, such as ICE, that would prevent information sharing. “We haven’t had any problems,” Katz said, “And we’re going to end the contract anyway because we always intended it to be a short-term solution.”

According to Katz, data analysis previously conducted with Palantir’s help will be brought in-house following the contract’s expiration.

Related

Alex Karp Insists Palantir Doesn’t Spy on Americans. Here’s What He’s Not Saying.

“Palantir makes money by enabling mass violence in the U.S. and around the world. They should have no place in our hospitals, our pension funds, or our government,” said Kenny Morris, an organizer with the American Friends Service Committee, which shared the contract documents with The Intercept.

“Our campaign against Palantir doesn’t stop in NYC,” Morris said. “We will continue to isolate this company and limit its destructive influence on our lives. In this city and around the world, communities are organizing to push more and more corporate clients, institutions, and politicians to cut ties with Palantir.”

The post Palantir Will No Longer Profit Off of New Yorkers’ Health Data appeared first on The Intercept.

In an era when companies known for e-commerce, social networks, and search engines have also become close collaborators with militaries, is bombing their servers fair game?

Three days after the U.S. and Israel began their joint bombardment, the Islamic Revolutionary Guard Corps launched kamikaze drone strikes against Amazon-owned data centers in the United Arab Emirates and Bahrain that provide an array of cloud computing services to customers throughout the Middle East. The impacts and subsequent fires “caused structural damage, disrupted power delivery to our infrastructure, and in some cases required fire suppression activities that resulted in additional water damage,” according to Amazon, resulting in service outages across the region.

The motive behind the attack, according to Iranian state television, was not to block people from ordering groceries or posting to social media, but rather to highlight “the role of these centers in supporting the enemy’s military and intelligence activities.” Though only Amazon’s centers are known to have come under fire, a March 11 tweet from the quasi-official Tasnim News Agency listed dozens of regional facilities, including data centers owned by Microsoft, Google and others, deemed “Enemy Technology Infrastructure” suitable for targeting.

It’s unclear if the Amazon data centers struck by Iranian drone strikes are used for military purposes or civilian purposes, or both. And it’s unknown if the attacks in any way hindered the militaries of the U.S., Israel, or their allies in the Gulf from using AI or other cloud-based services in their war efforts. But with Amazon, Google, and even Facebook parent company Meta are all eager partners of the Pentagon that augment the destructive power of the United States in Iran and elsewhere, server farms may now have the same status as factories building bombs and warplanes.

Scholars of international law and the laws of armed conflict say that when a military runs on the cloud, the cloud becomes a legal military target. But the cloud is an abstraction, not a physical site — a global network of millions of chips in servers spread across hundreds of massive buildings across the planet, servicing both civilian apps and state tools used to surveil and kill. Separating the former from the latter is an extremely difficult task.

“The legality turns on whether the specific facility, at the specific moment, is genuinely serving the military operations of a party to the conflict in a way that offers a concrete and definite advantage to the attacker,” explained León Castellanos-Jankiewicz, a lawyer with the Asser Institute for International and European Law in The Hague.

Sometimes the split between military and civilian use is straightforward. Microsoft, for example, helps run the Joint Warfighter Cloud Capability, which the Pentagon says provides it with “greater lethality.” This work involves the processing of classified data, which the government does not want commingling with civilian tech. Cloud computing services are generally offered via geographically distinct “regions,” each made up of many physical data centers. Customers typically select the region that is closest to them to minimize lag time. Microsoft’s US DoD Central and US DoD East regions are “reserved for exclusive [Department of Defense] use,” according to the company, and are serviced by data centers in Des Moines, Iowa, and Northern Virginia, respectively.

Amazon offers similar cloud regions exclusive for Pentagon use, though the location of these data centers is not public. Oracle, another JWCC provider, operates Pentagon-specific facilities in Chicago, Phoenix, and Virginia. Companies are understandably tight-lipped about where exactly on the map these facilities stand, in no small part because Iran, or any country at war with the U.S., would have reason to target them.

“A data center that is used solely or primarily for military applications is targetable,” said Ioannis Kalpouzos, an international law scholar and visiting professor at Harvard Law, “and a center that supports the Pentagon’s JWCC falls in that category.”

Related

AI’s Imperial Agenda

The march of data center construction has become a point of contention across the United States and around the world, with communities frequently — and sometimes successfully — rallying to block what they view as enormous resource-draining eyesores. But for those living in the widening shadow of data centers, planned or built, their status as military targets may be unsettling beyond concerns over water and energy consumption.

And as Defense Secretary Pete Hegseth aggressively shoehorns AI tools into the military wherever possible, the rapid expansion of data centers means the potential proliferation of legitimate military targets across the United States.

With comparisons between the destructive power of AI-augmented warfare and nuclear weaponry becoming more common, the ever-expanding network of American data centers may recreate Cold War anxieties around intercontinental ballistic missile, or ICBM, silo placement. The country’s nuclear launch capabilities were famously clustered in the relatively sparsely populated Upper Midwest, forming a so-called “nuclear sponge” that would draw Soviet nukes away from population centers and toward rural areas and farmland.

But the legal calculus around most data centers will be less clear. Google, for example, says the Pentagon uses both its general purpose public cloud and smaller specialized air-gapped networks that don’t touch the public internet, depending on the sensitivity of the data involved. Even cloud work involving Top Secret military data “can operate within Google’s trusted, secure, and managed data centers.” The company also sells modular mini-data centers for use closer to battlefields or bases.

These arrangements, shrouded in both military and trade secrecy, make it hard to assess whether a server is hosting a student’s homework or Air Force R&D, blurring the legality of attacking data centers that may host both. Google may have little control over how governments use its cloud tools; The Intercept has previously reported that Google executives worried internally they wouldn’t be able to tell how the Israeli military was deploying its cloud services.

“The practical challenge is that cloud infrastructure is often technically opaque, even to providers themselves,” Castellanos-Jankiewicz said. “The services a given data center supports may not be readily ascertainable from the outside or even inside, which complicates the attacker’s legal obligations considerably.”

Amazon and Google’s Project Nimbus similarly provides cloud computing services across the Israeli government, including both civilian agencies and the Ministry of Defense, along with state-owned weapons companies.

“The picture becomes more legally complex when a data center functions as a so-called ‘dual-use’ object,” simultaneously hosting military data or capabilities alongside civilian services,” Castellanos-Jankiewicz told The Intercept. “Once a facility is found to make an effective contribution to military action, the entire physical object can, under the dominant legal view, qualify as a military objective.”

The embrace of commercial cloud computing by the U.S. and others has muddled an already murky legal picture, Castellanos-Jankiewicz explained. “A military’s decision to store classified data or run AI-enabled military systems on commercial cloud infrastructure shared with civilian services could itself raise legal concerns — particularly if the commingling of military and civilian uses makes a strike more likely or increases the foreseeable harm to civilians when one occurs.”

Related

OpenAI on Surveillance and Autonomous Killings: You’re Going to Have to Trust Us

Determining whether a given data center can be legally attacked under international humanitarian law — itself comprised of various treaties that not every country adheres to — relies on a complex series of balancing tests that rarely produce concrete answers. To begin with, every object and person is generally presumed civilian and exempt from attack under this framework. Before launching a strike, a country is supposed to have a verifiable reason to believe a data center contributes to the enemy war effort, and reason to believe an attack will appreciably harm that effort. What “effectively contributes to military action” will, of course, be a source of disagreement.

Anthropic’s Claude large language model was reportedly used to accelerate American airstrikes against Iran; Claude, in turn, was built in part using 500,000 chips housed in an $11 billion Amazon data center in Indiana. If Claude is now arguably a weapon, is this Indiana site the data equivalent of a bomb factory? Kalpouzos, the Harvard Law visiting professor, told The Intercept it depends on the facts at the moment the bomb hits, not past usage. “If the facility is currently used in the training of the LLM that is used in the conduct of military operations — for example, by fine-tuning object classification or user-interaction features — then this could render it targetable,” he said.

In a recent article for Just Security, Klaudia Klonowska and Michael Schmitt said that the law calls for proportionality and restraint even against military targets. An attack against a data center that provided both military and civilian computing would need to be precise enough to destroy the former while minimizing harm to the latter, they argued. But international law may call for a degree of carefulness that militaries have little interest in. “If it were possible to attack only the area of the data center where servers hosting military data are located without destroying the entire center, the attacker would need to do so,” they wrote.

These requirements can be hard to observe in reality. The U.S. and Israel both tout the extreme precision of airstrikes that regularly slaughter civilians. And neither country, nor Iran, is a signatory to some of the relevant legal frameworks that make up the so-called “laws of armed conflict” in the first place.

Indiscriminate warfare practice by U.S. and Israel has also, ironically, been instrumental in reshaping how these laws are interpreted and effectively loosened. Throughout the Israeli genocide in Gaza, Israel’s military and the Pentagon both made clear it’s acceptable to destroy an apartment block or hospital if one first claims there is a genuine military target inside.

The second Trump administration in particular has been keen to more tightly integrate Silicon Valley into the global American killing apparatus, a plan to which the industry has shown itself to be largely amenable. Even after being thoroughly maligned by the administration following the collapse of its Pentagon deal over purported disagreements around safety guardrails, Anthropic CEO Dario Amodei issued a public statement making clear he still wanted in on military spending: “Anthropic has much more in common with the Department of War than we have differences. We both are committed to advancing US national security and defending the American people, and agree on the urgency of applying AI across the government.” That attitude, now commonplace across the tech sector, will see the further commingling of consumer tech and warfare both in the abstract and under sprawling data center rooftops across the country.

“These [data centers] are further melding military and civilian infrastructure,” said Kalpouzos, “and together with the increasingly permissive rules of engagement adopted by the U.S. and Israel, are potentially drawing in larger sectors of the economy and society in what is targeted and destroyed.”

The post Data Centers Are Military Targets Now appeared first on The Intercept.