A Senate committee that oversees the Pentagon has adopted a measure aimed at keeping people — not algorithms — accountable when force is used.
According to KJZZ, the amendment comes from Arizona Sen. Mark Kelly and is attached to the annual defense bill. Its goal is to require what the measure calls "ultimate human responsibility" in the use of force, a phrase intended to ensure that decisions involving artificial intelligence remain under human control.
The details reported so far are limited. KJZZ describes the committee adopting Kelly's measure, but the source items here do not spell out the exact text, how the requirement would be enforced, or when it might take effect. As an amendment to a larger defense bill, it would still need to move through the broader legislative process before becoming law.
The push reflects a wider conversation in defense circles about how much trust to place in automated systems. A related video from Military Embedded Systems frames the issue around what it calls a "Trust Layer" for military AI — shorthand for the safeguards and verification that defense technologists argue any battlefield AI would need before commanders could rely on it.
Why it matters: as militaries fold AI into more of their operations, lawmakers are signaling that a human — not a machine — should bear final responsibility for life-and-death decisions, and this amendment is an early attempt to write that principle into U.S. law.