Ai Policy

How advanced AI could lower barriers to biological harm, and the layered safeguards — model-level refusals, know-your-customer checks, access controls and policy — that security and governance leaders should understand.

A high-level, policy-focused look at AI in weapons development — autonomy, the dual-use problem, export controls and human-in-the-loop requirements — and the safeguards that keep meaningful human control over the use of force.