Anthropic filed two federal lawsuits Monday against the Trump administration, fighting back after the Pentagon formally labeled the Claude AI maker a "supply chain risk" — a designation historically reserved for foreign adversaries like Huawei and ZTE.
The suits, filed in the U.S. District Court for the Northern District of California and the U.S. Court of Appeals for the D.C. Circuit, call the blacklisting "unprecedented and unlawful." Anthropic says it's already losing money: federal contracts are being canceled and private-sector deals are in doubt, threatening "hundreds of millions of dollars in the near-term."
What the Pentagon Did — and Why It's Unusual
Last Thursday, the Pentagon formally issued the supply chain risk designation against Anthropic. The label now requires every defense contractor and vendor to certify that they don't use Claude in any work with the U.S. military. That's a sweeping ban — Claude powers AI tools used across dozens of federal agencies.
The supply chain risk label has almost never been applied to a U.S. company. It's the same tool the government used against Chinese telecom giants when it wanted to cut them out of American infrastructure. Applying it to a San Francisco AI startup is, by any measure, an extraordinary escalation.
President Trump amplified the pressure on Truth Social: "WE will decide the fate of our Country — NOT some out-of-control, Radical Left AI company run by people who have no idea what the real World is all about."
What Anthropic Is Actually Arguing
This isn't a fight about whether the government has to buy Anthropic's products. The company explicitly says it's not trying to force federal contracts. The argument is narrower and more pointed: the Pentagon can't punish a company for its speech.
Anthropic's position is that the blacklisting is retaliation for its public advocacy — specifically, the company's vocal opposition to using Claude for mass domestic surveillance and autonomous weapons systems. According to the filing, the government disagreed with those policy stances and used the supply chain risk designation as a cudgel.
That's a First Amendment argument, and it's a meaningful one. The complaint says the designation "attacks Anthropic's reputation and core First Amendment freedoms." The company is asking the court to vacate the designation entirely and grant a stay while the case proceeds.
How We Got Here
The backstory is a two-week escalation that started quietly. Anthropic was an early government AI partner, with Claude deployed across agencies as Washington scrambled to modernize its systems. Then talks broke down — specifically over how Claude could be used on the battlefield.
Anthropic refused to lift restrictions on autonomous weapons and domestic mass surveillance. The Trump administration, deep in a war footing over the Iran conflict, pushed back hard. By late February, Trump posted his directive telling federal agencies to "immediately cease" all use of Anthropic's technology. The formal supply chain risk designation followed days later.
Amazon, which has committed up to $4 billion to Anthropic and sells Claude through AWS Bedrock to federal customers, is now watching a key partner fight for its government business. The blacklist ripples well beyond Anthropic's direct contracts.
Why This Matters Beyond Anthropic
The case sets a precedent that every AI company should be watching. If the Pentagon can blacklist a firm because it advocates for usage restrictions on its own technology, that's a chilling signal: build AI guardrails, lose government business.
OpenAI reversed course on some of its own military restrictions earlier this year. Whether that was coincidence or calculated positioning is unclear. But the contrast with Anthropic — which doubled down — is now stark.
There's also a broader market signal. Anthropic's safety-first positioning has been a selling point for enterprise customers. A federal supply chain risk label, even if eventually reversed, damages that reputation in sectors where government trust matters: healthcare, defense primes, financial services.
The outcome will shape how AI labs navigate the tension between safety principles and government access for years. If Anthropic wins, it establishes that AI companies can set usage limits without being punished by the state. If it loses, the message to every lab is clear: compliance is the price of admission.
FOLLOW NeuralWire on X for daily AI signal — what matters, why it matters, what to do about it. →