Artificial Intelligence Risk Evaluation Act of 2025

• Authority: United States Congress
• Parent authority: Federal government
• Jurisdiction: United States
• Collections: U.S. federal laws
• Most recent activity: Proposed
• Most recent activity date: 2025-09-29

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
• Cyberattacks, weapon development or use
• Competitive dynamics
• Governance failure
• AI pursuing its own goals
• AI possessing dangerous capabilities
• Lack of capability or robustness

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

• Disinformation, surveillance, and influence at scale
• Fraud, scams, and targeted manipulation
• Loss of human agency and autonomy
• Increased inequality and decline in employment
• Lack of transparency or interpretability

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), malicious actors (4.1, 4.2), competitive dynamics (6.4), governance failure (6.5), and AI safety failures (7.1, 7.2, 7.3). Coverage is concentrated in security, misuse prevention, governance, and AI safety domains, with particular emphasis on loss-of-control scenarios and existential risks.

Sector coverage

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

• Information
• Scientific Research & Development Services
• National Security

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

• Trade, Transportation & Utilities

All other sectors had no coverage in this document.

Sectors Summary

This is a cross-sectoral federal regulation that applies to all covered advanced AI system developers regardless of their industry sector. The document does not limit governance to specific economic sectors but rather regulates AI developers based on the compute threshold used to train their systems. The Act has implications for critical infrastructure and national security across multiple sectors.

Legislative Status

Hard Law

Legislative Status Summary

This is a binding federal statute (bill) introduced in the U.S. Senate that establishes mandatory obligations for covered AI system developers, includes specific penalties for non-compliance ($1,000,000 per day), and creates enforceable prohibitions on deployment without compliance.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Entities: United States Congress; Senator Hawley; Senator Blumenthal
• Summary: The bill was introduced in the U.S. Senate by Mr. Hawley (for himself and Mr. Blumenthal) on September 29, 2025, and referred to the Committee on Commerce, Science, and Transportation.

Target

• Entity Types: AI_Developer
• Entities: covered advanced artificial intelligence system developers
• Summary: The Act explicitly targets 'covered advanced artificial intelligence system developers' defined as persons that design, code, produce, own, or substantially modify advanced AI systems trained using computing power greater than 10^26 operations.

Enforcer

• Entity Types: AI_Governance_Actor
• Entities: Secretary of Energy; Department of Energy
• Summary: The Secretary of Energy is designated as the primary enforcement authority, responsible for establishing and administering the Advanced Artificial Intelligence Evaluation Program, with authority to impose fines for violations.

Monitor

• Entity Types: AI_Governance_Actor
• Entities: Secretary of Energy; Department of Energy; Congress; independent third-party assessors
• Summary: The Secretary of Energy monitors through the testing program, with oversight by Congress through required reports. The program also facilitates independent third-party assessments for transparency.

AI Lifecycle Stages

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

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

AI Lifecycle Summary

The document primarily focuses on the Verify and Validate, Deploy, and Operate and Monitor stages. It establishes comprehensive testing and evaluation requirements before deployment, mandatory participation in evaluation programs, and ongoing monitoring of AI systems. There is minimal coverage of earlier stages like planning, data collection, or model building.

Technical Scope

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

• AI Models
• AI Systems
• Compute Threshold
• Open Weight / Open Source

Technical Scope Summary

The document explicitly defines and covers AI systems and advanced AI systems with a specific compute threshold of 10^26 operations. It extensively addresses artificial superintelligence and discusses various AI capabilities. The document does not explicitly mention foundation models, generative AI, predictive AI, general purpose AI, task-specific AI, or open-weight models, though it does reference 'open-source advanced artificial intelligence system' in the deployment definition.