Anthropic CEO denies Pentagon's request to eliminate AI safeguards

Anthropic has stated it will not relent in its dispute with the US Department of Defense (DoD) regarding the application of its artificial intelligence (AI) technology.

Firm's leader Dario Amodei remarked on Thursday that they would opt out of collaborations with the Pentagon rather than consent to uses of their tech that "threaten, not protect, democratic principles."

His statement follows a recent discussion with US Secretary of Defense Pete Hegseth about demands for Anthropic to accept "any legal use" of its tools. 

The meeting wrapped with a threat of potentially removing Anthropic from the DoD's supply network.

"These threats do not alter our stance: we cannot in good faith comply with their demand," Amodei declared.

The primary concern for Anthropic revolves around its AI tools such as Claude being utilised for "Widespread domestic monitoring" and "Completely independent weapons systems."

Amodei noted "such applications have never been part of our agreements with the Department of War, and we believe they are still inappropriate."

The Department of War, a secondary title for the Defense Department, emerged under an executive order issued by US President Donald Trump in September.

"If the Department decides to end collaboration with Anthropic, we are prepared to facilitate a seamless transition to an alternative provider," Amodei stated.

A Pentagon official previously conveyed to the BBC that if Anthropic does not comply, Hegseth would ensure the Defense Production Act is enforced on the company.

The act effectively grants a US president the power to designate a particular company or its product as vital, compelling it to meet national defense needs.

However, Hegseth also warned of identifying Anthropic as a "supply chain threat," implying the company would be deemed insufficiently secure for government usage.