United States - AI Governance Act (H.R. 7532)

Overview

H.R. 7532, the Federal A.I. Governance and Transparency Act of 2024, creates a statutory Subchapter IV in chapter 35 of title 44, U.S. Code, establishing binding requirements for Federal artificial intelligence systems. The bill centralizes agency governance obligations and places the Office of Management and Budget (OMB) in the lead role for issuing government-wide guidance in consultation with the National Institute of Standards and Technology (NIST), the General Services Administration (GSA), the Office of Science and Technology Policy (OSTP), and other relevant agencies. Core deliverables include agency AI Governance Charters, a public Federal AI System Inventory maintained through a GSA interface, mandatory risk assessment and mitigation, privacy and civil rights protections, and independent evaluation and biennial reporting by Inspectors General and the Government Accountability Office (GAO). For the official bill text see Congress.gov — H.R. 7532 Text and for the Committee report see GovInfo — House Report 118-905.

Definitions

The bill provides operational definitions intended to clarify the scope of coverage and to avoid ambiguity: "artificial intelligence system" encompasses any data system, software, application, tool, or utility that uses static or dynamic machine learning algorithms or other forms of AI, whether as a primary purpose or embedded capability. "Federal artificial intelligence system" is an AI system used in connection with a Federal information system. The bill adopts standard definitions for "Federal information system" (per title 40) and "national security system" (per title 44) and includes definitions for governance artifacts such as "AI Governance Charter" and "Federal AI System Inventory" to create consistent obligations across executive branch agencies.

Governance and Institutional Framework

OMB is the central coordinating authority under the bill: it must develop, coordinate, and oversee policies, standards, and guidance to ensure appropriate use of Federal AI systems and protection of civil rights, civil liberties, and privacy (including conformity with the Privacy Act at 5 U.S.C. 552a). The Director of OMB will be required to consult NIST (for technical standards), GSA (for inventory and public interface), OSTP, agency CIOs, Chief Data Officers, and other relevant heads. Agencies must delegate primary authority for compliance to their Chief Information Officer (or equivalent) and integrate AI governance into strategic, data, workforce, and budgeting plans. The bill further specifies that OMB guidance should promote risk-based safeguards, removal of barriers to responsible use (IT, data, workforce, budget), and alignment with NIST standards referenced in the William M. (Mac) Thornberry National Defense Authorization Act for FY2021. See OMB and NIST coordination described in the reported text at Congress.gov — H.R. 7532 Text and the Committee's rationale in House Report 118-905.

Key Focus Areas

The statute concentrates on several interlocking policy domains: transparency and public notice, risk management and safety, privacy and civil rights protections, procurement and contractor obligations, independent validation and oversight, and consolidation of existing statutory requirements. Specific focus includes (1) publication of an AI Governance Charter for each Federal AI system (including purpose, use cases, testing and validation methods, responsible officials, maintenance plans, and impacted data assets), (2) establishment of a centrally maintained Federal AI System Inventory with public access via a GSA-managed interface, (3) integration of risk assessment and mitigation into acquisition and operational processes, (4) preservation of appeals and human review rights for program determinations influenced by AI, (5) protection of training data, algorithms, and system components from misuse or adversarial manipulation, and (6) requirements that agency contracts contain clauses enabling agencies to obtain information from contractors needed for compliance. The bill emphasizes that transparency is conditioned on protecting classified and otherwise exempt information under FOIA and other law; in those cases, charters may be submitted to OMB, congressional committees, and Inspectors General in lieu of public posting.

Implementation Framework

H.R. 7532 imposes a phased implementation approach: OMB must issue government-wide guidance within one year of enactment, in consultation with NIST, GSA, OSTP, and others. Within six months of OMB guidance, the Federal Acquisition Regulation (FAR) must be updated to require contractor disclosures necessary for agency compliance. Agencies must prepare and publicly post plans (using an OMB-provided template) to align with the subchapter's requirements, designate accountable officials, and maintain charters on agency webpages and the Federal AI System Inventory. The bill includes procedures for limited waivers and non-public filings when national security or other legal protections apply; these require advance OMB approval and notification to relevant congressional committees and agency Inspectors General. The implementation design is intended to balance transparency and security while enabling procurement and technical modernization to support safe AI adoption.

Monitoring and Evaluation

To ensure effectiveness and continuous improvement, H.R. 7532 requires independent evaluation and periodic reporting. Agency Inspectors General and the GAO will provide biennial assessments on agency compliance and the overall effectiveness of Federal AI governance. The Director (OMB) will oversee agency compliance and may use enforcement authority available under title 40 to ensure accountability. The Federal AI System Inventory provides auditable, centralized publication of governance charters enabling public and congressional oversight. These monitoring mechanisms are meant to identify systemic vulnerabilities, recurring failures in testing/validation, and areas where additional technical standards or legislative action may be necessary.

Penalties, Liability, and Appeals

The bill itself does not establish novel criminal penalties; instead enforcement focuses on administrative oversight, procurement remedies, and statutory authorities referenced in title 40. OMB oversight may include administrative actions to compel agency compliance, and updated FAR clauses provide contractual levers against noncompliant vendors. The statute preserves individuals' existing rights to appeal determinations and requires agencies to ensure alternative review processes independent of AI outputs. Liability and redress remain governed by existing tort, statutory, and administrative law; H.R. 7532 strengthens transparency and documentation requirements which may influence liability analyses but does not create an independent private right of action. Inspectors General reviews and GAO reports provide additional remedies through congressional oversight and possible legislative follow-up.

Relationship to Other Instruments

H.R. 7532 consolidates and streamlines certain provisions of the AI in Government Act of 2020 and the Advancing American AI Act of 2022 by recodifying overlapping requirements into the new Subchapter IV and repealing redundant language. It explicitly aligns with existing privacy law (Privacy Act, 5 U.S.C. 552a) and references technical standards promulgated under the National Institute of Standards and Technology (NIST) and the William M. (Mac) Thornberry National Defense Authorization Act for FY2021. The bill also directs updates to the FAR and coordinates with applicable FOIA exemptions and national security statutes for handling classified implementations. The Committee report explains these harmonization measures and the intent to reduce contradictory or duplicative agency obligations; see House Report 118-905.

International Alignment

The Act is narrowly domestic in scope (targeting Federal agency AI use) but acknowledges the value of aligning with international standards and best practices through references to NIST standards and interagency consultation. By codifying documentation, transparency, and third-party evaluation practices, H.R. 7532 can facilitate cross-border cooperation and interoperability with partner frameworks (e.g., OECD AI principles, ISO standards) without directly regulating private-sector exports or foreign entities. The statute's emphasis on standards-based technical guidance and risk-based safeguards supports future alignment with multilateral approaches to AI governance.

Implementation Timeline

Sources and References