Sponsor
Mark Edward Kelly
Senator · D-AZ
Senator · D-AZ
The bill meaningfully increases safety, accountability, training, and oversight of autonomous weapons—reducing unintended harm and legal risk—while imposing higher costs, added bureaucracy, operational constraints in time‑critical scenarios, and potential security/transparency trade‑offs.
Military personnel and civilians in conflict zones will face fewer unintended engagements and civilian casualties because the bill requires safety testing, red-team assessments, incident reporting, and mitigation policies before and after fielding.
Commanders and operators gain clearer legal and operational definitions of 'autonomous weapon system' and 'ultimate human responsibility', improving accountability and reducing wrongful autonomous uses of force.
Congress, oversight bodies, and commanders will get better records, required reporting, and post‑engagement review data (logs, decision logic, and recommendations), strengthening accountability and enabling corrective policy or legislative action.
Taxpayers and defense programs will face higher costs and longer timelines because of stricter design requirements, expanded testing, red-team assessments, logging, retraining, reporting, and a new office and repository.
Requiring clear human‑in‑the‑loop/ultimate responsibility and intervention capability may constrain system automation and responsiveness, reducing effectiveness in high‑speed or time‑critical engagements.
Detailed logging, publication of unclassified standards, and centralized repositories could expose sensitive decision logic, vulnerabilities, or testing methodologies if not tightly protected, creating operational security risks.
Based on analysis of 9 sections of legislative text.
Directs DoD to ensure autonomous weapon systems preserve human commanders/operators' ultimate responsibility via design, testing, training, reporting, and oversight.
Official title: Establish requirements for human judgment in the use of force by autonomous weapon systems used by the Department of Defense, and for other purposes.
Introduced June 4, 2026 by Mark Edward Kelly · Last progress June 4, 2026
Requires the Department of Defense to design, test, operate, train for, and report on autonomous weapon systems and AI capabilities so human commanders and operators retain "ultimate human responsibility" for use of force. It sets technical, operational, training, testing, reporting, and incident‑reporting requirements, creates a civilian harm mitigation office inside the Office of the Secretary of Defense, and directs public testing standards and semiannual reporting to Congress. Establishes independent testing and red‑team certification before fielding systems that can independently select and engage targets; mandates operator training pipelines, recordkeeping for post‑engagement review, a centralized incident repository modeled on aviation safety systems, and protections for operator reporting. Preserves the use of human‑supervised intercept systems for incoming threats.