This bill, the Responsible Artificial Intelligence for Veterans Act of 2026, requires the Secretary of Veterans Affairs to seek an agreement with a federally funded research and development center to do an independent review of artificial intelligence (AI) systems used by the Veterans Health Administration (VHA). The Secretary must seek the agreement within 90 days of the bill becoming law. The review must prioritize at least five AI systems that are deployed at scale or pose elevated clinical, operational, or patient safety risks. The evaluator must report results one year after the agreement, and the report must be unclassified and unredacted (a classified annex is allowed). The review must cover integration and readiness, governance and monitoring, model performance and bias, safety and human oversight, veteran-centered design, costs, documentation, scalability, and environmental effects where relevant. The Secretary must respond to the review within 120 days with a corrective action plan assigning responsibilities, milestones, and any remaining risks. The Comptroller General must review the evaluation and the corrective action plan and brief congressional veterans committees within 180 days after the plan is submitted.
If you are a veteran, clinician, or staff at VHA, the bill means an outside research center will examine AI tools used for care and operations. The review may identify safety, privacy, or performance problems and recommend fixes, including possible suspension of deployment. The bill does not itself order immediate changes to care; it requires evaluation, a departmental response, and follow-up review by the Comptroller General.
The bill says the Secretary must carry out the work using amounts the Department already has available and that no additional funds are authorized. No publicly available information on specific cost estimates or budget effects is provided in the bill text.
The bill directs an independent, technical review to find risks, improve governance, increase transparency, and protect veteran safety when AI is used in clinical settings. It requires timelines for reporting, a departmental corrective-action plan, and oversight by the Comptroller General and congressional veterans committees.
No publicly available information.