Sponsors (6)
The bill trades stronger federal oversight, mandatory risk assessments, independent verification, and emergency powers to reduce catastrophic AI risk and increase transparency against significant compliance costs, concentrated enforcement authority, possible secrecy around evidence, and legal uncertainty that could slow innovation and limit state-level remedies.
The American public and national-security stakeholders: large frontier model developers must assess, mitigate, and (for very large models) subject catastrophic-risk controls to independent, recurring assessments and audits, reducing the chance of AI-caused mass harm or existential-scale failures.
Frontier AI developers and regulators: the Secretary gains an expedited emergency authority to suspend or restrict development or deployment of models that pose imminent catastrophic risk, enabling rapid intervention to prevent immediate harms.
Developers, policymakers, and the public: new transparency requirements (published risk-mitigation frameworks, annual reviews, public reports, and audit findings) plus required reporting to Congress improve oversight and public accountability of frontier AI safety practices.
Large and very large frontier developers (and ultimately consumers): the combined requirements for assessments, licensed IVOs, recurring audits, reporting, recordkeeping, and secure handling create substantial compliance and operational costs that can raise product prices, slow deployments, and favor incumbents over startups.
Developers, researchers, and investors: broad enforcement powers (including rapid suspension authority) plus steep civil and criminal penalties and high daily fines create risk of abrupt business disruption and may chill innovation or risky-but-beneficial R&D.
Developers, third parties, and the public: mandatory access to detailed (sometimes unredacted) materials for auditors/regulators together with FOIA exemptions concentrates information with the federal government while raising IP, privacy, and confidentiality concerns and limiting public scrutiny.
Based on analysis of 9 sections of legislative text.
Creates a federal safety regime: licensed independent verifiers, mandatory public safety frameworks and audits for large frontier AI developers, emergency authority to restrict models posing imminent catastrophic risk, and preempts certain state rules.
Official title: To provide for Federal oversight of the development and deployment of frontier artificial intelligence in interstate and foreign commerce, and for other purposes.
Introduced July 23, 2026 by Jay Obernolte · Last progress July 23, 2026
Requires licensing and oversight of independent verification organizations (IVOs), forces very large and large frontier AI developers to publish and follow public safety frameworks, mandates periodic third-party assessments, and gives a federal official emergency authority to suspend or restrict AI development or deployment that poses an imminent catastrophic risk. Sets definitions and thresholds for "catastrophic risk," creates licensing, auditing, reporting, and market-assessment duties, and preempts most new state rules on developer obligations in three specified areas.