Overview
- Sen. Adam Schiff introduced the Human Authority in Lethal Operations (HALO) Act on June 8, proposing a law that would force a human commander to have final authority for any lethal action involving AI.
- The bill would require the Pentagon to name commanders who authorize AI-assisted strikes and keep detailed records of how targets are selected to enable post-action review and accountability.
- The HALO Act would bar the use of covered AI to monitor people engaged in constitutionally protected activities and explicitly forbid removing humans from decisionmaking tied to presidential nuclear actions.
- Senators Elissa Slotkin and Kirsten Gillibrand have filed similar measures and sponsors are seeking to press these rules into the annual NDAA, even as Pentagon officials say existing military doctrine already preserves human authority.
- The push follows a public split with AI firm Anthropic over Pentagon access to models and could force formal reviews, remediation, or retirement of fielded systems if Congress enacts statutory guardrails.