Sponsor
Kirsten Gillibrand
Senator · D-NY
Senator · D-NY
The bill strengthens DoD oversight, human control, and incident reporting for high‑risk military AI—improving safety and accountability—but does so at the cost of significant compliance costs, potential delays to urgent deployments, and renewed civil‑liberty and transparency risks if carve‑outs or classified processes are overbroad.
DoD and commanders will get faster, clearer incident reporting, definitions, and reporting pathways so threats from compromised or dangerous AI models are detected and contained more quickly.
Service members and operators gain stronger pre-deployment safety checks, human authority to pause/terminate or override AI actions, and required training and fail‑safes, reducing risk of catastrophic or unintended AI-enabled actions.
Specific civil‑liberty protections: U.S. persons in the United States are protected from government AI-enabled identity resolution, tracking, and automated scoring without a lawful predicate.
Contractors and DoD programs face substantial new compliance, certification, training, and documentation burdens that will raise costs, strain small vendors, and increase procurement prices for taxpayers.
Tight pre-deployment certification, notification, congressional approval rules, and related human‑in‑the‑loop requirements risk slowing or blocking time‑sensitive or defensive AI deployments in crises.
Broad definitions and DoD roles in domestic person‑based analysis, plus carve-outs and classified authorizations, risk expanded surveillance and erosion of privacy and civil liberties for Americans.
Based on analysis of 8 sections of legislative text.
Imposes DoD rules: define AI incident/reporting, designate high-consequence AI, require senior certification and human accountability, ban certain AI uses, mandate contractor reporting, and require DFARS updates.
Official title: Provide for secure and accountable use of artificial intelligence by the Department of Defense, and for other purposes.
Introduced June 2, 2026 by Kirsten Gillibrand · Last progress June 2, 2026
Creates new Defense Department rules for use, procurement, reporting, and oversight of advanced AI systems used by the military. It defines terms, requires DoD to designate "high-consequence" AI categories, mandates human accountability and certification before deployment, imposes contractor incident-reporting requirements, bans certain AI uses (including AI-directed nuclear targeting and broad domestic person-based analytics), sets limits and exception procedures for autonomous weapons, updates acquisition rules, and requires congressional review for any otherwise-prohibited autonomous weapon systems. Sets deadlines for DoD rulemaking and certification processes (many within 30–180 days of enactment), requires rapid contractor reporting of security incidents and risky model behaviors, and creates an expedited congressional approval path for specific autonomous weapon systems that are not otherwise allowed under the law.