NEED TO KNOW

• A federal judge said Tuesday the Pentagon's Anthropic blacklist looks like punishment for public criticism of military AI use
• The supply-chain risk designation — historically reserved for foreign adversaries — has never been applied to a U.S. company before
• If the injunction fails, contractors including Amazon, Microsoft, and Palantir must certify they don't use Claude

SAN FRANCISCO, CA (TDR) — A federal judge sharply questioned the Pentagon's legal basis for blacklisting Anthropic on Tuesday, suggesting the designation looks designed to punish the AI company for speaking publicly about its safety limits — not to protect national security.

The big picture: The standoff between Anthropic and the Trump administration has moved from social media to a San Francisco courtroom, and the outcome could define how the government exercises power over AI companies that set limits on military use of their products.

Freedom-Loving Beachwear by Red Beach Nation - Save 10% With Code RVM10

• The Pentagon designated Anthropic a supply-chain risk to national security in early March — a label historically applied to foreign firms like Huawei
• The designation prohibits Defense Department agencies and all contractors from doing business with the company
• Anthropic has a second lawsuit pending in Washington, D.C., over a separate designation that could extend the blacklist to civilian federal contracts

Why it matters: The case is no longer just about one AI contract — it's a test of whether the executive branch can use national security law to compel private AI companies to accept terms they've rejected on safety grounds.

• Defense contractors including Amazon, Microsoft, and Palantir would be required to certify non-use of Claude if the designation stands
• AI researchers and legal scholars warn a precedent forcing companies to accept "all lawful uses" removes the only institutional check on autonomous weapons deployment
• Anthropic says the designation will cost the company billions in lost government and commercial contracts

Driving the news: Tuesday's hearing in the U.S. District Court for the Northern District of California centered on Anthropic's request for a preliminary injunction to pause enforcement while the full lawsuit is litigated.

CLICK HERE TO READ MORE FROM THE THE DUPREE REPORT

• U.S. District Judge Rita Lin questioned whether the three Trump administration actions — Trump's ban, Hegseth's contractor requirement, and the supply-chain designation — were legally tailored to any genuine national security concern
• Lin noted the Pentagon could simply stop using Claude without extending the blacklist to all federal contractors and their commercial relationships
• The judge called the government's argument that its social media announcements weren't legally binding "pretty surprising," given that they are central to the lawsuit
• Lin said she expects to issue a ruling on the injunction within days

What they're saying: The hearing surfaced a core disagreement over whether Anthropic's refusal to accept contract terms was a policy dispute or protected speech.

• Anthropic lead counsel Michael Mongan — "This is something that has never been done with respect to an American company. It is a very narrow authority. It doesn't apply here."
• Mongan argued that the designation's sweep — forcing contractors to cut all commercial ties with Anthropic — goes far beyond what the supply-chain risk statute permits
• DOJ attorney Eric Hamilton, speaking for the government — the Pentagon "came to worry that Anthropic may in the future take action to sabotage or subvert IT systems"
• Hamilton argued the designation was based on contractual disagreement, not Anthropic's public statements on AI safety
• ACLU Deputy Director Patrick Toomey — "Anthropic's public advocacy for AI guardrails is laudable and protected by the First Amendment — not something the Pentagon should be punishing"

Yes, but: Anthropic's position carries its own complications the company hasn't fully answered publicly.

• The Pentagon argues that Anthropic retains operational control over Claude even when deployed in classified settings — a claim Anthropic denies but hasn't resolved with technical transparency
• The government's core concern — that Anthropic could update Claude in ways that disrupt military systems — reflects a real tension in any AI contract where the developer controls future model behavior
• By refusing "all lawful uses," Anthropic effectively claimed a veto over Pentagon operations, which courts may view differently than a simple speech act

Between the lines: The Trump administration's escalation pattern — social media announcement, formal designation, contractor ban — suggests the goal was market pressure, not security remediation.

• Legal experts note that Hegseth's X post announcing the contractor prohibition exceeded the statute's scope before any formal determination was made — a procedural gap the government is now trying to paper over in filings
• The Pentagon has pivoted to OpenAI and xAI, both of which reportedly accepted "all lawful uses" terms — signaling the blacklist functions as a loyalty test for the broader AI industry, not a one-off dispute

What's next:

• Judge Lin expected to rule on the preliminary injunction within days
• If granted, Anthropic may continue federal and contractor work while litigation proceeds
• Parallel lawsuit in the D.C. Circuit addresses a second designation potentially covering civilian agency contracts
• Sen. Elizabeth Warren has sent a formal letter to Hegseth demanding justification for the blacklisting

If the government can designate a domestic AI company a national security risk for setting safety limits on its own product — what stops the same authority from being used against any firm that declines a federal contract on ethical grounds?

Sources

This report was compiled using information from NPR, CNBC, Axios, Al Jazeera, MarketScreener, Android Headlines, and Euronews.

Freedom-Loving Beachwear by Red Beach Nation - Save 10% With Code RVM10