Anthropic PBC announced it is holding 'productive conversations' with defense officials concerning its artificial intelligence safety feud with the Pentagon. Chief Executive Dario Amodei stated the company is working with officials to find ways to continue collaborating with the Pentagon while adhering to AI safety policies.
Last June, Anthropic secured a contract to provide the Pentagon with access to its AI models. The company previously stated it would not permit its software for domestic surveillance or the development of fully autonomous weapons. This led to an order from President Donald Trump for federal agencies to cease using Claude.
More recently, U.S. Defense Secretary Pete Hegseth instructed the Pentagon to designate Anthropic as a supply chain risk, a move that limits how military suppliers can use the company's products. Amodei noted this designation is not expected to significantly impact the majority of Claude users.
Anthropic plans to legally challenge the supply chain risk designation, calling it 'legally unsound and set a dangerous precedent.' The company is also working with officials on a potential transition should Claude be phased out of public sector networks, as federal agencies have been given six months to switch AI providers.