OMB Issues Draft AI Risk Management Guidance for Federal Agencies — August 29, 2023
The Office of Management and Budget released draft policy instructing U.S. federal agencies to strengthen AI governance, risk management, and procurement controls consistent with the AI Bill of Rights and NIST frameworks.
The draft memorandum would require agencies to appoint chief AI officers, implement risk management playbooks for safety-impacting systems, and release public inventories of AI use cases. It also proposes minimum evaluation and impact assessment practices aligned with the Blueprint for an AI Bill of Rights.
- Governance structure. Agencies must stand up AI governance boards, designate responsible executives, and integrate civil rights officials in reviews.
- Risk controls. The policy outlines testing, monitoring, incident response, and human oversight standards for safety-impacting AI.
- Transparency. Agencies would publish inventories and provide opt-out processes when AI affects the public.
Federal program leads used the consultation period to align upcoming budgets and contracts with the proposed compliance benchmarks.