Anthropic AI poses 'unacceptable risk' to military, US government says

WASHINGTON: The US government on Tuesday maintained that artificial intelligence firm Anthropic posed an "unacceptable risk" to military supply chains, as it defended its decision to sever ties with the company against a legal challenge.

Anthropic's Claude AI model has drawn attention in recent weeks over its alleged use in identifying targets for US bombing operations in Iran, as well as the company's refusal to allow its systems to be deployed for mass surveillance within the United States or to power lethal fully autonomous weapons systems.

Pentagon defends cutting ties with Anthropic

In a court document seen by AFP, the Pentagon — which has been rebranded as the Department of War (DoW) by the Trump administration — justified its decision to cut ties with Anthropic in response to a legal complaint filed by the firm.

"Became concerned that allowing Anthropic continued access to DoW's technical and operational warfighting infrastructure would introduce unacceptable risk into DoW supply chains."

The government further argued in its filing to a California court that AI systems are particularly susceptible to interference.

"AI systems are acutely vulnerable to manipulation."

Anthropic's legal challenge

The filing came as the Pentagon defended itself against Anthropic's legal challenge to its designation as a security risk. The AI company had lodged a legal complaint contesting the government's move to end its relationship with the firm.

The case underscores the growing tension between the US defence establishment and prominent AI companies over the use of artificial intelligence in military applications. Anthropic has been at the centre of this debate, having publicly taken a position against certain military uses of its technology, including its opposition to powering lethal fully autonomous weapons systems and mass surveillance programmes.

The dispute highlights the broader questions surrounding the role of AI in modern warfare and the extent to which technology companies can resist government demands for the use of their products in military and intelligence operations. The Trump administration's rebranding of the Pentagon as the Department of War adds a further dimension to the evolving relationship between the defence establishment and the technology sector.

The legal battle between Anthropic and the US government is being closely watched as it could set important precedents for how AI companies interact with military and government agencies going forward.