Bloomberg Talks Pentagon’s Emil Michael on Anthropic Talks, Military’s AI Use
Feb 27, 2026
Emil Michael, U.S. Undersecretary of Defense for Research and Engineering, who leads Pentagon research and tech policy, talks about recent talks with Anthropic and concessions offered. He discusses negotiations, minor differences that stalled progress, options like the Defense Production Act, and the role of autonomous systems and other AI vendors for classified use.
AI Snips
Chapters
Transcript
Episode notes
Negotiations Were Closer Than Public Messages Indicated
- Negotiations between the Pentagon and Anthropic were near final despite public claims of a breakdown.
- Emil Michael says the Department offered written concessions and was surprised when Anthropic publicly cut off talks.
Require Legal Guarantees And Human Oversight In AI Contracts
- Do include explicit legal and oversight assurances when contracting AI with the military.
- Michael says the DoD agreed in writing to follow laws like the National Security Act and FISA and to require human oversight for autonomous systems.
DoD Has Multiple Levers If Talks Fail
- The Pentagon sees multiple legal and policy tools if talks fail, including the Defense Production Act or supply-chain risk designations.
- Michael says the final course of action by Secretary Hegseth depends on the outcome at the deadline and remains under consideration.