Unpacking the Great American AI Act

On June 4, 2026, Representative Jay Obernolte (R-CA) and Representative Lori Trahan (D-MA) released a discussion draft of artificial intelligence (AI) legislation titled the “Great American AI Act” (GAAIA). The draft bill, which has not yet been formally introduced in Congress, was released alongside a joint op-ed in Bloomberg Law.

Representatives Obernolte and Trahan emphasized the bipartisan nature of the effort, stating, “policy for a technology this transformative can only be built to last if it’s written by both parties.”

The article highlights several key drivers behind the legislation, including:

• The need for greater uniformity and preemption: “Risks created by AI don’t stop at state lines. The most advanced systems are built in one state and used in all 50, shaping jobs, consumers, and public safety everywhere. Protections that depend on your zip code are not enough.”
  
  
• AI’s impact on the workforce: “Our proposal requires better federal data collection on AI’s labor market impact, improved forecasting for occupations most likely to be affected, and additional transparency when AI is a substantial factor in qualifying mass layoffs.”
  
  
• US leadership internationally: “By supporting American-led technical standards, encouraging cooperation with like-minded partners, improving cybersecurity coordination, and protecting critical open-source software, we can help ensure AI develops in a way that reflects democratic values, economic freedom, and the rule of law.”

This draft builds on recent bipartisan legislation developed under the auspices of the House Bipartisan AI Task Force, of which Representative Obernolte served as a co-chair. Representative Trahan also served on the task force.

Representatives Trahan and Obernolte also sit on the House Energy and Commerce Committee, which is the committee of jurisdiction for various AI- and tech-related matters.

While several members have joined in supporting the discussion draft, including Representatives Scott Franklin (R-FL), Suhas Subramanyam (D-VA), Erin Houchin (R-IN), and Scott Peters (D-CA), the House Democratic Commission on AI came out in opposition hours after release of the draft text.

The draft legislation spans 269 pages and contains titles devoted to:

• Frontier AI Government (Title I)
• Workforce (Title II)
• Cybersecurity (Title III)
• Research, Development, and International Cooperation (Title IV)

Transparency in frontier AI models

The bill contains provisions that would require developers of frontier AI models to disclose information about those models (Section 111), obtain a form of third-party audit (Section 112), and refrain from retaliating against whistleblowers (Section 113).

Section 111

Section 111 would require large frontier AI model developers to publish a “frontier AI framework” focused on risk mitigation and the handling of “catastrophic risk.” All frontier model developers would be required to publish transparency reports containing additional information and submit reports of any “critical safety incidents” to the Director of the Center for Artificial Intelligence Standards and Innovation (CAISI).

Section 112

Section 112 would create a federal system in which the CAISI Director would license and oversee “independent verification organizations (IVOs)” that large frontier model developers would have to retain to perform “audits and assessments.”

The IVOs would verify developer compliance with the law and assess the adequacy of the developer’s “frontier AI framework, governance policies and practices, risk monitoring, and mitigation of detected risks for achieving acceptable levels of mitigation of catastrophic risks.”

The IVOs would perform these audits and assessments semi-annually, and on an ad hoc basis at the CAISI Director’s request, and they would provide the Director with post-audit reports. They could also refer violations to the Attorney General (AG) and state AGs. Such referrals would be mandatory in matters involving imminent catastrophic risk.

Violating the requirements in Sections 111 and 112 would subject developers to civil penalties in actions brought by the AG or, under certain circumstances, by state AGs.

Both Sections 111 and 112 would sunset in three years.

Section 113

The whistleblower provision, Section 113, protects employees and independent contractors against retaliation for providing information regarding “AI violations.” This term is defined to mean any violation of federal law or regulations “related to the development, deployment, or operation of artificial intelligence.”

Similarities to other recently enacted bills

All three sections borrow heavily from recently enacted frontier model laws in California and New York, as well as a similar Illinois law that Governor JB Pritzker is expected to sign soon. The Illinois law is the only one of the three to require third-party audits and is thus the closest parallel to the IVO provisions of the federal bill.

Another recently enacted state law in Virginia focuses on IVOs but imposes no requirements, instead directing state officials to evaluate the feasibility and impact of an IVO framework.

Notable preemption provisions

Title I of the bill would preempt state laws and rules “specifically regulating the development of” any AI model, where “development” is broadly defined as “acts performed or directed by a developer prior to its deployment.” The bill explains that the purpose of such preemption is to allow the federal government to “establish testing, evaluation, and oversight frameworks.” Such preemption would last for three years unless reauthorized.

The bill makes clear that it does not apply post-deployment or preempt state laws or rules of "general applicability." Thus, many AI-related state bills and laws would remain unaffected, such as laws regulating companion bots and commercial chatbot interactions.

However, the preemption language could preempt several state laws imposing obligations on developers, including the three frontier developer laws described above.

GAAIA incorporates multiple bipartisan bills

The draft text incorporates provisions from several bipartisan bills on AI policy addressing a variety of issues, including cybersecurity, AI education and workforce, fraud deterrence, whistleblower protections, data center energy efficiencies, and others.

For instance, the draft bill incorporates text from a longstanding bipartisan measure, the Future of Artificial Intelligence Innovation Act of 2026, that seeks to codify the establishment of CAISI with a director appointed by the Secretary of Commerce, who would oversee evaluations and assessments of AI systems in the US and competitors developing AI systems overseas. It also includes recent legislation that would extend key cybersecurity information sharing authorities with liability protections to incentivize parties to share cyber threat information and strengthen cyber defenses.[1] These authorities and protections are set to expire at the end of the fiscal year if Congress does not act.

Below is a list of bipartisan bills also incorporated into GAAIA:

• S. 3952: Future of Artificial Intelligence Innovation Act of 2026
  • Sponsor: Senator Todd Young (R-IN)
    
    
• H.R. 10281: Promoting United States Leadership in Standards Act of 2024
  • Sponsor: Representative Jay Obernolte (R-CA-23)
    
    
• H.R. 6461: Resources for Evaluating and Documenting AI Models (READ) AI Models Act
  • Sponsor: Representative Sarah McBride (D-DE-At Large)
    
    
• H.R. 7434: AI Grand Challenges Act of 2026
  • Sponsor: Representative Ted Lieu (D-CA-36)
    
    
• H.R. 2385: Creating Resources for Every American To Experiment (CREATE) with AI Act of 2025
  • Sponsor: Representative Jay Obernolte (R-CA-23)
    
    
• H.R. 5332: Liquid Cooling for AI Act of 2025
  • Sponsor: Representative Jay Obernolte (R-CA-23)
    
    
• H.R. 5584: Literacy in Future Technologies (LIFT) AI Act
  • Sponsor: Representative Thomas H. Kean (R-NJ-7)
    
    
• H.R. 9403: Expanding AI Voices Act
  • Sponsor: Representative Valerie P. Foushee (D-NC-4)
    
    
• H.R. 5351: National Science Foundation (NSF) AI Education Act of 2025
  • Sponsor: Representative Vince Fong (R-CA-20)
    
    
• H.R. 3460: AI Whistleblower Protection Act
  • Sponsor: Representative Jay Obernolte (R-CA-23)
    
    
• H.R. 6306: AI Fraud Deterrence Act
  • Sponsor: Representative Ted Lieu (D-CA-36)
    
    
• H.R. 5079: Widespread Information Management for the Welfare of Infrastructure and Government (WIMWIG) Act
  • Sponsor: Representative Andrew R. Garbarino (R-NY-2)

Key takeaways

In releasing the discussion draft, the Congressional sponsors are seeking feedback from stakeholders, experts, and the public before the bill is formally introduced. Interested parties have an opportunity to comment.

Learn more

For more information, please contact the authors.

 

^{[1] H.R. 5079, the Widespread Information Management for the Welfare of Infrastructure and Government (WIMWIG) Act, introduced by Representative Andrew R. Garbarino (R-NY).}