Artificial Intelligence Research, Innovation, and Accountability Act of 2023

• Authority: United States Congress
• Parent authority: Federal government
• Jurisdiction: United States
• Collections: U.S. federal laws
• Most recent activity: Defunct
• Most recent activity date: 2025-01-03

Risk subdomain coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.

• AI system security vulnerabilities
• False or misleading information
• Pollution of information ecosystem
• Disinformation, surveillance, and influence at scale
• Fraud, scams, and targeted manipulation
• Governance failure
• Lack of capability or robustness
• Lack of transparency or interpretability

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.

• Unfair discrimination and misrepresentation
• Unequal performance across groups
• Compromise of privacy
• Overreliance and unsafe use
• AI possessing dangerous capabilities

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has good coverage of approximately 10-12 subdomains, with strong focus on AI system security (2.2), misinformation (3.1, 3.2), malicious actors (4.1, 4.3), governance failure (6.5), and AI safety failures (7.3, 7.4). Coverage is concentrated in security, transparency, accountability, and governance domains.

Sector coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.

• Information
• Finance & Insurance
• Real Estate, Rental & Leasing
• Educational Services
• Health Care & Social Assistance
• Public Administration (excl. National Security)

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.

• Trade, Transportation & Utilities
• Professional & Technical Services
• Scientific Research & Development Services
• National Security

All other sectors had no coverage in this document.

Sectors Summary

The document primarily governs AI use across multiple high-impact sectors including Finance and Insurance, Health Care, Educational Services, Real Estate, Professional Services, Public Administration, National Security, and Information sectors. The governance is comprehensive and cross-sectoral, focusing on AI systems that make decisions affecting housing, employment, credit, education, healthcare, insurance, criminal justice, and critical infrastructure.

Legislative Status

Hard Law

Legislative Status Summary

This is a binding federal statute enacted by Congress with mandatory obligations, enforcement mechanisms including civil penalties up to $300,000, and explicit prohibition authority. The document uses mandatory language throughout ('shall', 'must', 'required') and establishes clear enforcement procedures through the Secretary of Commerce and Attorney General.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Entities: United States Congress; Senate; House of Representatives
• Summary: The document is proposed federal legislation introduced by the United States Congress, as indicated by the opening text 'Be it enacted by the Senate and House of Representatives of the United States of America in Congress assembled'.

Target

• Entity Types: AI_Developer, AI_Deployer, AI_Infrastructure_Provider
• Entities: covered internet platforms; deployers of high-impact artificial intelligence systems; critical-impact AI organizations; developers of critical-impact artificial intelligence systems
• Summary: The Act explicitly targets multiple categories of AI actors including developers and deployers of AI systems, particularly those operating high-impact and critical-impact AI systems, as well as covered internet platforms using generative AI.

Enforcer

• Entity Types: AI_Governance_Actor
• Entities: Secretary of Commerce; Under Secretary of Commerce for Standards and Technology; Director of the National Institute of Standards and Technology; Attorney General; Office of Management and Budget; Federal Trade Commission
• Summary: The Act designates the Secretary of Commerce as the primary enforcement authority with power to impose penalties, issue orders, and prohibit AI system deployment. The Attorney General has authority to bring civil actions, and various other federal agencies have oversight roles.

Monitor

• Entity Types: AI_Governance_Actor
• Entities: National Institute of Standards and Technology (NIST); Secretary of Commerce; Comptroller General of the United States; Office of Management and Budget; Artificial Intelligence Certification Advisory Committee
• Summary: NIST is designated to develop standards and conduct research, the Secretary reviews compliance reports and assessments, the Comptroller General conducts reviews of AI use in government, and an advisory committee is established to oversee certification standards.

AI Lifecycle Stages

Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems

• Plan and Design
• Collect and Process Data
• Build and Use Model
• Verify and Validate
• Deploy
• Operate and Monitor

AI Lifecycle Summary

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 design requirements, data collection through transparency reporting, model development through developer obligations, extensive validation through TEVV standards, deployment through certification requirements, and ongoing monitoring through biennial assessments and transparency reports.

Technical Scope

Terms relating to technical scope that are mentioned or implicitly referenced in the document

• AI Models
• AI Systems
• Generative AI
• Predictive AI

Technical Scope Summary

The document explicitly mentions AI systems and AI models throughout, with detailed definitions provided. It addresses generative AI extensively through specific transparency requirements. The document does not explicitly mention frontier AI, general purpose AI, task-specific AI, foundation models, predictive AI, or open-weight models. There are no compute thresholds specified.