Ensuring Safe and Ethical AI Development Through SAFE AI Research Grants

HR 6402 establishes a federal grant program administered by the National Academy of Sciences to fund research on developing "safe AI models" and mitigating AI risks. The program requires the Director to create public guiding principles for AI safety through stakeholder input, then submit a detailed proposal within one year outlining grant structure, research priorities, and evaluation methods. It directly affects researchers and institutions seeking federal funding for AI safety studies, not end-users or specific AI products. Key provisions include mandatory public input on ethical guidelines, a requirement to assess existing AI models' safety features, and a timeline for implementing the grant program. The bill focuses on research funding, not regulation of AI technology or deployment.