Official name: Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence
Mandates appointment of a Chief AI Officer within 60 days. Requires development of AI strategies. Imposes risk management requirements. Obligates annual AI use case inventory. Encourages removing innovation barriers while ensuring compliance with federal policies.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding OMB memorandum with mandatory requirements, enforcement mechanisms, and legal authority under multiple federal statutes including the AI in Government Act of 2020 and Executive Order 14110. It uses mandatory language throughout and establishes specific compliance deadlines with consequences for non-compliance.
The document has good coverage of approximately 10-12 subdomains, with strong focus on AI system security (2.2), governance failure (6.5), lack of robustness (7.3), lack of transparency (7.4), unfair discrimination (1.1), privacy compromise (2.1), and false information (3.1). Coverage is concentrated in governance, system safety, discrimination prevention, and risk management domains.
This document governs AI use across all federal government operations, which means it applies to AI systems used in Public Administration (excluding National Security for certain provisions) and National Security contexts. The document also indirectly governs AI use across multiple sectors through federal agency operations including Healthcare, Education, Law Enforcement (within Public Administration), Financial Services (federal programs), and Transportation (federal systems). The governance is sector-agnostic in its core requirements but includes sector-specific use cases in Appendix I.
The document comprehensively covers all AI lifecycle stages with particular emphasis on Verify and Validate, Deploy, and Operate and Monitor stages. It addresses planning through AI strategies, data collection and processing requirements, model development practices, extensive testing and validation requirements, deployment procedures, and ongoing monitoring obligations.
The document explicitly mentions AI models, AI systems, and generative AI with detailed definitions and requirements. It references dual-use foundation models and model weights. It does not explicitly define or distinguish frontier AI, general purpose AI, task-specific AI, predictive AI, or open-weight models, though it addresses concepts related to these categories. No specific compute thresholds are mentioned.
The memorandum is issued by OMB (Shalanda D. Young, Director) under authority of multiple federal statutes and executive orders. OMB is the primary proposer with coordination from OSTP.
OMB has primary enforcement authority through reporting requirements, compliance reviews, and approval of extensions/waivers. CAIOs have delegated enforcement authority within their agencies. Authorizing officials enforce through Authorization to Operate processes.
OMB monitors through annual inventories, certifications, and reporting requirements. CAIOs monitor agency AI use cases and compliance. AI Governance Boards provide internal oversight. An interagency council coordinates across agencies.
The memorandum explicitly applies to federal agencies that develop, use, or procure AI. CFO Act agencies have additional requirements. Intelligence Community elements are excluded from certain sections.
15 subdomains (8 Good, 7 Minimal)