Comply or Die: The Pentagon Blacklisted an American AI Company for Having Ethics
Weekly Deep Dive
The Pentagon told an American company to drop its ethical guardrails. The company said no. So the government treated it like a hostile foreign power.
Last Friday, Defense Secretary Pete Hegseth designated Anthropic a “Supply-Chain Risk to National Security,” a classification historically reserved for foreign adversaries like Huawei and Kaspersky. Anthropic’s crime: refusing to let the military use its AI for mass domestic surveillance and fully autonomous weapons without any restrictions.
Hours later, OpenAI swooped in with a Pentagon deal for classified networks, claiming its agreement includes the same safeguards Anthropic got blacklisted for requesting. According to several officials, this is a lie and the Pentagon actually has free-reign to do whatever they want now with OpenAI’s systems.
The Shakedown
The setup started months ago. Anthropic signed a $200 million contract with the Pentagon in July 2025, becoming the first frontier AI company to deploy models on classified military networks. The company asked for two guardrails: no mass domestic surveillance of Americans, and no fully autonomous weapons. The Pentagon wanted “all lawful purposes” with zero exceptions.
On February 24, Hegseth issued an ultimatum: agree to the Pentagon’s terms by 5:01 PM Friday, February 27, or face consequences. The government also threatened to invoke the Korean War-era Defense Production Act, a 1950 law that lets the military commandeer private company facilities.
CEO Dario Amodei didn’t blink: “We will not knowingly provide a product that puts America’s warfighters and civilians at risk.” He offered to work with the Pentagon on R&D to improve autonomous system reliability. The Pentagon wasn’t interested.
At 5:01, the deadline passed. Trump took to Truth Social: “The Leftwing nut jobs at Anthropic have made a DISASTROUS MISTAKE trying to STRONG-ARM the Department of War.” He ordered every federal agency to immediately cease using Anthropic’s technology, with a six-month phase-out.
Hegseth’s follow-up was even more aggressive: “Effective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic.”
That’s not canceling a vendor contract. That’s economic warfare against a $380 billion American company with $14 billion in annualized revenue that just raised $30 billion and was planning an IPO.
Follow the Money, Follow the Musk
Four days before Anthropic got blacklisted, Elon Musk’s xAI signed its own deal to put Grok on classified Pentagon systems. Musk, Trump’s biggest financial backer in 2024, has been publicly attacking Anthropic, claiming the company “hates Western civilization.”
Small problem: government insiders told the Wall Street Journal they consider Grok too sycophantic, too susceptible to data poisoning, and less capable than Claude. The GSA, the agency that handles federal procurement, flagged security concerns. Until Anthropic refused the Pentagon’s ultimatum, military officials preferred Claude.
So the government sanctioned the AI it trusted, to make room for the AI it doesn’t, owned by the president’s biggest donor. As the CSIS’s Greg Allen put it: “The Defense Department just sent a huge message to every company that if you dip your toe in the defense contracting waters, we will grab your ankle and pull you all the way in, anytime we want.”
The Legal House of Cards
The supply chain risk designation under 10 USC 3252 is designed to protect against adversaries who might “sabotage, maliciously introduce unwanted function, or otherwise subvert” military systems. Anthropic isn’t sabotaging anything. It’s setting terms on a product it built. Multiple legal experts say the designation is on shaky ground.
The statute requires a risk assessment and Congressional notification before the designation takes effect. It’s unclear whether either happened. Hegseth’s claim that every military contractor must cut all commercial ties with Anthropic goes far beyond what the statute allows, according to the company and several federal contracting lawyers.
Dean Ball, a senior fellow at the Foundation for American Innovation and a former Trump administration AI policy adviser, didn’t mince words. He called it “the most shocking, damaging, and overreaching thing I have ever seen the United States government do” and described Hegseth’s interpretation as “almost surely illegal,” “attempted corporate murder,” and “obviously a psychotic power grab.”
That’s not a progressive critic. That’s a former Trump administration AI policy adviser.
The Dominoes Are Still Falling
Anthropic says it will challenge the designation in court. The legal fight will test whether the government can weaponize national security classifications against domestic companies that refuse to surrender their terms of service. If Amazon, Google, and Nvidia, all Anthropic investors who also do business with the Pentagon, actually have to divest, the chilling effect on every tech company in America would be immediate and permanent.
The precedent here isn’t about AI. It’s about whether the government can economically destroy any company that negotiates instead of capitulates. If a $380 billion American firm with $14 billion in annualized revenue can get treated like Huawei for asking that its technology not be used for domestic mass surveillance, what chance does anyone else have?
Sources:
- CNBC: Trump admin blacklists Anthropic
- NPR: OpenAI announces Pentagon deal after Trump bans Anthropic
- WIRED: Anthropic Hits Back After ‘Supply Chain Risk’ Label
- Anthropic: Statement on comments from Secretary of War
- OpenAI: Our agreement with the Department of War
- Fortune: OpenAI sweeps in after Anthropic labeled ‘supply chain risk’
- Forbes: Hegseth designates Anthropic as supply chain risk
- Futurism: Government insiders concerned by Grok deployment
Awful! Thank you for this article!