Senator Adam Schiff (D-Calif.) introduced legislation Monday that would require human involvement whenever the Pentagon deploys autonomous or semi-autonomous weapons systems that rely on artificial intelligence, according to reporting from MSN. The bill is part of a broader push by Democratic senators to write AI guardrails directly into the National Defense Authorization Act, the annual legislation that sets military policy and spending, according to The Hill.
Axios, citing an exclusive, described the measure as a bill for "responsible" defense AI. The Debrief reported that lawmakers framed the effort in urgent terms, warning of a "potential threat" and calling for Congress to "act now" to prevent AI from making life-and-death military decisions without a human in the loop.
Not everyone sees the legislation as a hard stop. Responsible Statecraft noted that lawmakers are pumping the brakes on military AI "sort of" — suggesting the bill may carve out exceptions or leave room for considerable AI autonomy in practice.
The U.S. military has been accelerating its use of AI for logistics, surveillance, and targeting support, raising long-standing questions among ethicists, arms-control advocates, and some members of Congress about accountability when machines recommend — or take — lethal action.
If passed, the law would mark one of the first binding statutory limits on how the American military can use AI in combat contexts, setting a precedent for how democracies govern lethal autonomy at a moment when the technology is advancing faster than the rules designed to contain it.