Hegseth gives Anthropic Friday deadline on military AI
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Pete Hegseth threatened to remove Anthropic from the Pentagon's supply chain at a Tuesday meeting with CEO Dario Amodei if the company declined to allow its AI to be used across military applications.
- Anthropic has until Friday evening to comply, according to a senior Pentagon official, with the Pentagon threatening to invoke the Defense Production Act and label the company a supply chain risk if it does not.
- Anthropic's stated red lines include AI involvement in autonomous kinetic operations where tools make final military targeting decisions without humans, and use of its technology for mass domestic surveillance.
- The Pentagon disputes Anthropic's framing, with a senior official telling the BBC the conflict is unrelated to autonomous weapons or mass surveillance and that the company should have no say in how its products are used.
- The dispute stems from a reported breach of trust after Anthropic's Claude was used in the January operation that led to the capture of former Venezuelan President Nicolás Maduro, via a contract with Palantir.
- Anthropic, OpenAI, Google, and xAI each received Pentagon contracts of up to $200m (£148m) last summer, though the Pentagon says only Anthropic is resisting broad military deployment of its models.
Why it matters: Anthropic was the first tech company cleared for the Pentagon's classified military networks, and forced compliance via the Defense Production Act would override the company's stated ethical guardrails on its $200m defense contract — setting a precedent for how much say AI labs retain when the US government deems their restrictions a supply chain risk.
