Emil Michael explains why US flags Claude AI as security risk

The US Defence Department has formally designated Anthropic’s Claude AI models as a national security supply-chain risk, marking the first time an American AI company has received this classification.

In an interview with CNBC on Squawk Box, Defence Department Chief Technology Officer Emil Michael stated that “different policy preferences” embedded in its design could impact the Pentagon’s supply chain.

“We can’t have a company with policy preferences baked into the model pollute the supply chain so our warfighters receive ineffective weapons, body armour or protection,” Michael said.

He further stated that the designation was not made with the intention of punishing Anthropic, as only a fraction of their revenue came from US government contracts.

Why Anthropic is challenging decision?

Anthropic has filed a lawsuit against the Trump administration, stating that the move is “unprecedented and unlawful” and that hundreds of millions of dollars in contracts are at risk.

The issue started when the Pentagon asked Anthropic to remove the limits that prevent their artificial intelligence system from being used for fully autonomous weapons or domestic surveillance. The company declined to comply due to reliability issues and the possible infringement of human rights.

When negotiations did not work out, Defence Secretary Pete Hegseth issued the national security supply chain risk designation. The move was also ordered by Trump himself to implement a six-month phase-out of existing contracts.

The Pentagon argued that the law regarding artificial intelligence must be followed, not private policies. The limits set by Anthropic posed a threat to national safety.

Michael acknowledged that their artificial intelligence system cannot be deleted or replaced immediately. “This is not just Outlook, where you could delete it from your desktop,” he said.