Great American AI Act Freezes State AI Laws Three Years
Key insights
- The bill freezes new state AI development laws for three years while preserving state authority over deployment, labor, and consumer privacy.
- Frontier AI developers with over $500M revenue face semi-annual independent audits and $1 million per day penalties for non-compliance.
- The bill authorizes $100 million per year for fiscal years 2027-2029 to fund the Center for AI Standards and Innovation (CAISI).
Why this matters
State AI preemption at the federal level would halt California's AB 2013 and portions of SB 942, targeting development-side regulations specifically while leaving deployment and consumer privacy laws intact. Frontier AI developers including Anthropic, OpenAI, xAI, and Google DeepMind would face a new compliance regime with semi-annual independent audits and $1 million per day penalties, creating significant legal and operational overhead before a certified auditor ecosystem even exists. With the AFL-CIO, the House Commission on AI, and consumer advocates already opposing the draft, the bill's path to passage tests whether federal AI governance can satisfy both labor and industry constituencies simultaneously.
Summary
A bipartisan House draft released June 4 would freeze new state AI development laws for three years, prompting immediate opposition from labor groups and congressional critics.
The 269-page proposal from Reps. Obernolte and Trahan targets frontier AI developers with over $500M annual revenue for semi-annual audits, with penalties up to $1 million per day for non-compliance.
Essentially: Anthropic, OpenAI, xAI, and Google DeepMind face the audit regime.
- The AFL-CIO called it a "hard no"; the House Commission on AI stated it does not support the draft as currently written.
- Safety incidents require reporting within 15 days, or 24 hours for imminent threats.
The Information Technology Industry Council backed it as a national standard, widening the industry-vs-consumer divide.
Potential risks and opportunities
Risks
- Anthropic, OpenAI, xAI, and Google DeepMind face compliance uncertainty for semi-annual CAISI audits before a certified auditor ecosystem exists, potentially exposing them to $1 million per day penalties while standards are still being defined.
- California's AB 2013 and portions of SB 942 face a three-year development-side freeze, suspending state-level protections in those areas while the federal framework is built out.
- Consumers lose development-side AI protections in states that had enacted such laws, with no federal replacement mechanism explicitly filling that gap during the three-year preemption period.
Opportunities
- Compliance and audit firms capable of obtaining CAISI certification could see immediate demand from Anthropic, OpenAI, xAI, and Google DeepMind before the framework is even finalized.
- The Information Technology Industry Council and aligned trade groups gain first-mover leverage to shape CAISI audit standards by being the principal institutional backers of the national framework.
- AI safety incident monitoring vendors benefit from mandatory 15-day reporting requirements (24 hours for imminent threats) that will drive demand for automated compliance infrastructure at frontier developers.
What we don't know yet
- Which organizations will qualify as CAISI-certified auditors and who oversees their accreditation — the discussion draft establishes the requirement but not the infrastructure to fulfill it.
- Whether California's AB 2013 and the affected portions of SB 942 would be grandfathered or immediately suspended upon enactment — not clarified in the draft.
- What specific changes the AFL-CIO and House Commission on AI require before they could support the bill — neither organization specified terms in their opposition statements.
Originally reported by techtimes.com
Read the original article →Original headline: Great American AI Act Would Freeze State AI Laws for Three Years, Imposing Semi-Annual Audits on Frontier Developers — AFL-CIO Calls It a 'Hard No'