Secretary of War Tweets That Anthropic is Now a Supply Chain Risk
A rogue tweet threatens to ban all DoD contractors from using Anthropic's AI, wiping $150B from markets.
In a dramatic escalation of tensions between Anthropic and the U.S. Department of Defense, Secretary of War Pete Hegseth issued a rogue tweet declaring Anthropic a 'Supply Chain Risk.' This move came just an hour after President Trump had announced a six-month wind-down of Anthropic's government contract, which was seen as a de-escalation. Hegseth's tweet demanded that all DoD contractors cease using any Anthropic products 'effective immediately,' a directive with a potentially massive global blast radius. The trigger was Anthropic's refusal to allow its AI models, like Claude, to be used for analyzing bulk data collected from Americans—including search history, GPS movements, and financial transactions—for Pentagon intelligence purposes.
Hegseth's unilateral declaration, described as 'illegal' and 'corporate murder' by observers, wiped approximately $150 billion from public markets in post-close trading. The situation highlights a stark divergence in AI ethics and government contracting: while Anthropic held its red line on bulk data analysis, OpenAI accepted a DoD contract on the same day, reportedly agreeing to similar terms. The incident underscores the immense power and risk concentrated in executive agencies regarding critical technology supply chains and sets a chilling precedent for how AI policy disputes might be resolved outside formal legal and diplomatic channels.
- Secretary Hegseth's unilateral tweet declared Anthropic a 'Supply Chain Risk,' demanding all DoD contractors stop using its AI immediately.
- The crisis was triggered by Anthropic's refusal to let its AI analyze bulk American user data (searches, location, transactions) for the Pentagon.
- The announcement caused a $150B market drop and contrasts with OpenAI's acceptance of a similar DoD contract on the same day.
Why It Matters
Sets a dangerous precedent for executive power over AI supply chains and highlights the ethical divide between major AI labs on government surveillance.