OMB Publishes Draft Guidance Implementing Procedural Requirements for Agencies Developing, Using, and Procuring AI

In response to the Biden-Harris Administration’s recent Executive Order 14110 entitled “Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence,” the Office of Management and Budget (OMB) today released draft guidance outlining federal agencies’ obligations and suggested actions around the responsible development, use, and procurement of AI technologies.

The OMB guidance implements three broad changes to federal agency practices involving AI. First, it establishes new agency roles and resources for managing government AI systems, including a new Chief AI Officer role that must lead each agency’s implementation of the OMB guidance; new agency AI Governance Bodies comprised of senior agency officials that meet quarterly to discuss each agency’s AI governance procedures; and cross-agency AI Use Case Inventories accompanied by public AI use case reporting.

Second, the OMB guidance requires agencies to build internal processes, strategies, and capacity to increase their use and procurement of responsible AI. These internal processes and strategies are meant to increase flexibility for and remove barriers to AI development and use through investment in IT infrastructure, incorporation of public and commercial databases, and increased hiring and training for agency employees in AI positions.

Third, the OMB guidance sets out minimum AI risk management practices that most executive agencies are expected to follow when developing, using, or procuring AI systems that impact individuals’ rights or safety. These practices include ongoing AI impact assessments covering an AI system’s intended purpose, potential risks, and relevant data; real-world performance testing to ensure reliability and risk mitigation in practice; independent evaluations of AI performance; annual AI monitoring for model drift and degradation; human review of AI decision-making; and consultations with affected and underserved communities.

For years, EPIC has been a leading advocate for stronger oversight of government AI systems, including efforts to ban government use of biometric surveillance tools and push agencies to adopt the Universal Guidelines for A.I. and the OECD A.I. Principles. EPIC has also published extensive research on current AI risks, including reports on generative AI and government AI procurement. EPIC looks forward to working with federal agencies like OMB to carry out Executive Order 14110 and ensure that AI systems reflect our fundamental values and serve the public interest.