Anthropic has promised a legal challenge after the Pentagon labelled it a security risk and banned it from government contracts, cranking up its row with the Trump administration.
US president Donald Trump gave the start-up six months until it is cut from government work, accusing it of making a “disastrous mistake” by challenging the military use of Claude.
US president Donald Trump: “ALLOW A RADICAL LEFT, WOKE COMPANY TO DICTATE HOW OUR GREAT MILITARY FIGHTS AND WINS WARS!” He said he was “directing EVERY Federal Agency in the United States Government to IMMEDIATELY CEASE all use of Anthropic’s technology”.
Trump offered a “six-month phase-out period for Agencies like the Department of War who are using Anthropic’s products, at various levels”, leaving the door ajar for a last-minute stitch-up. Two people familiar with the matter said the Pentagon was still negotiating with Anthropic right up to the moment Trump hit post.
The Pentagon then designated Anthropic a supply-chain risk, a move usually saved for foreign firms and hostile jurisdictions. Groups previously hit with that label include China’s Huawei and ZTE Corporation and Russia’s Kaspersky Lab.
OpenAI, chief executive Sam Altman, announced a deal to deploy models on the government’s classified networks, with Elon Musk’s xAI said to be close to its own deal. Defence Secretary Pete Hegseth went for the throat, saying Anthropic had “delivered a masterclass in arrogance and betrayal” after a deadline lapsed on Friday.
Defence Secretary Pete Hegseth added: “may conduct any commercial activity with Anthropic”. That “effective immediately” ban was aimed at Pentagon contractors and suppliers, turning a policy spat into a procurement bloodbath.
Anthropic said it would take the government to court. Anthropic: “No amount of intimidation or punishment from the Department of War will change our position on mass domestic surveillance or fully autonomous weapons. We will challenge any supply chain risk designation in court.”
Anthropic argued Hegseth did not have “the statutory authority to back up this statement and said a supply-chain risk designation can only extend to the use of Claude as part of Department of War contracts”. It said the Pentagon cannot dictate how contractors use Claude for other customers.
University of Minnesota Law School, associate professor Alan Rozenshtein said, “the supply chain risk designation is legally tenuous. I wouldn’t be surprised if all sides, including the government, understand this”.
Anthropic, chief executive Dario Amodei has clashed with Hegseth over red lines on lethal autonomous weapons and mass domestic surveillance, refusing to budge. On Thursday, he said his company “cannot in good conscience” agree to the US government’s terms.
For now, Claude is still the only AI model deployed in classified operations, having been used in the seizure of Venezuelan leader Nicolás Maduro last month. An administration official told the FT it remained the best model for military use, while US senator Mark Warner said Trump’s move posed “an enormous risk to US defence readiness”.
