Pentagon chief technology officer Emil Michael stated that there is currently no resolution in sight between the Pentagon and Anthropic, despite the Pentagon recently securing agreements with other frontier AI companies that allow unrestricted access, excluding Anthropic. This ongoing standoff arose when Anthropic refused Pentagon demands to disable safety features on their Claude models for use in applications such as autonomous weapons. The Department of War is also shifting towards a diversification strategy to reduce reliance on any single AI provider, a move prompted by prior over-reliance issues.

Anthropic: Anthropic is an AI safety and research company dedicated to building reliable, interpretable, and steerable AI systems such as its Claude models. It recently addressed code quality issues in Claude and launched initiatives exploring AI’s economic and labor market impacts. In this news, Anthropic faces an ongoing dispute with the Pentagon, designated as a supply chain risk for refusing to remove safety guardrails limiting military use of its models.
Emil Michael: Emil Michael serves as Under Secretary of War for Research and Engineering and Chief Technology Officer for the Department of War, leading efforts to integrate advanced technologies like AI into defense operations. Formerly a top Uber executive, he oversees negotiations with AI firms and prioritizes diversification in AI sourcing. In this context, he stated there is no resolution in sight with Anthropic despite new pacts with rival AI providers.

Recent Deals: Pentagon secured agreements with other frontier AI companies allowing unrestricted access, excluding Anthropic.
Dispute Origin: The standoff began when Anthropic declined Pentagon demands to disable safety features on Claude models for applications including autonomous weapons.
Diversification Shift: Department of War is actively moving away from dependence on any single AI provider after prior over-reliance.