Judge says government’s Anthropic ban looks like punishment

A federal choose in San Francisco mentioned on Tuesday the federal government’s ban on Anthropic seemed like punishment after the AI firm went public with its dispute with the Pentagon over the navy’s potential makes use of of its synthetic intelligence mannequin, Claude.

U.S. District Decide Rita F. Lin made the comment on the outset of a listening to about Anthropic’s request for a preliminary injunction in one in all its lawsuits towards the Pentagon, which has designated the corporate a provide chain threat, successfully blacklisting it.

“It seems to be like an try and cripple Anthropic,” Lin mentioned, including she was involved that the federal government is likely to be punishing Anthropic for brazenly criticizing the federal government’s place.

Lin mentioned she anticipated to make a ruling within the subsequent few days on whether or not to briefly pause the federal government’s ban till the courtroom decides on the deserves of the case.

The listening to within the U.S. District Courtroom for the Northern District of California is the newest improvement in a spat between one of many main AI firms and the Trump administration, and it has implications for a way the federal government can use AI extra broadly.

Anthropic CEO Dario Amodei introduced in late February that he wouldn’t enable the corporate’s Claude’s AI mannequin for use for autonomous weapons, or to surveil Americans. President Trump subsequently ordered all U.S. authorities businesses to cease utilizing Anthropic’s merchandise.

The Pentagon designated Anthropic as a “provide chain threat” earlier this month, citing nationwide safety issues. That designation is generally reserved for entities deemed to be international adversaries that might doubtlessly sabotage U.S. pursuits.

Anthropic has filed two federal lawsuits alleging that this designation quantities to unlawful retaliation towards the corporate for its stance on AI security. It argues that the label will price it each clients and income, since it can bar Pentagon contractors from doing enterprise with the corporate, as effectively.

The lawsuits, filed within the U.S. District Courtroom for the Northern District of California and the federal appeals courtroom in Washington, D.C., allege the Trump administration violated the corporate’s First Modification proper to speech and exceeded the scope of provide chain threat legislation.

In at present’s listening to, legal professionals for Anthropic mentioned it was apparently the primary time such a designation had been made towards a U.S. firm.

Lin mentioned the Pentagon has a proper to resolve what AI merchandise it desires to make use of. However she questioned whether or not the federal government broke the legislation when it banned its businesses from utilizing Anthropic, and when Protection Secretary Pete Hegseth introduced that anybody searching for enterprise with the Pentagon should minimize relations with Anthropic.

She mentioned the actions have been “troubling” as a result of they didn’t appear to be tailor-made to the nationwide safety issues in query, which could possibly be addressed by the Pentagon merely ceasing to make use of Claude. As a substitute, she mentioned, it seemed like the federal government was attempting to punish Anthropic.

However a lawyer for the federal government argued that its actions weren’t retaliatory, and have been primarily based on Anthropic’s disagreement with the federal government over how its AI mannequin could possibly be used — not the corporate’s choice to talk out about it.

The federal government additionally argued that Anthropic is a threat as a result of, theoretically, sooner or later the corporate may replace Claude in a method that endangers nationwide safety.

Anthropic didn’t reply instantly to an emailed request for remark.

A Pentagon spokesperson mentioned that the company’s coverage is to not touch upon ongoing litigation.