On June 4, 2026, two members of Congress released what may be the most significant federal AI governance proposal in US history. Representatives Jay Obernolte (R-CA) and Lori Trahan (D-MA) published a 269-page bipartisan discussion draft called the Great American Artificial Intelligence Act, and it is worth understanding what it contains — not because it is law yet, but because it signals where federal AI governance is heading.

What Is in the Bill

The draft is organised into four major sections: Frontier AI Governance, Workforce, Cybersecurity, and Research, Development, and International Cooperation. The parts with the most immediate relevance to businesses are the first two.

Frontier AI Governance creates a new compliance framework for what the bill calls “large frontier developers” — companies with more than $500 million in annual revenue that have trained a frontier AI model. If you fall into that category, the bill would require you to:

• Publicly disclose key information about your frontier models
• Submit to third-party audits of those models
• Protect AI whistleblowers from retaliation

For most businesses using AI as customers of OpenAI, Anthropic, Google, or Microsoft rather than as model trainers, these obligations do not directly apply. The governance burden falls on the model makers, not the model users.

Workforce provisions do touch a wider range of companies. The draft would require employers to give 60 days advance notice when AI is a “substantial factor” in a mass layoff. That notice would need to include information about the AI system involved and an estimate of how many job losses are attributable to it. If you are planning to automate operations at scale, this is a new disclosure obligation worth tracking.

The State Preemption Question

The most politically charged element of the draft is a three-year preemption of state AI laws. If enacted, this would freeze state-level AI regulation for three years. The Colorado AI Act, California’s AI Transparency Act, and similar state rules would be put on hold while federal standards develop.

This is a notable shift from earlier proposals that called for a 10-year freeze. By limiting it to three years, the bill’s sponsors are trying to make the trade-off more politically workable. States give up immediate enforcement power, and in return they get a federal framework rather than a permanent preemption.

Whether this is a good deal depends on your perspective. Businesses managing compliance across multiple states would welcome three years of federal consolidation. Consumer advocates and state regulators are less enthusiastic, arguing it strips protections that are already in place.

The preemption is not yet law, and the draft still needs to survive the formal legislative process before it becomes anything binding. But the direction of travel is clear.

Still a Discussion Draft

It is worth being precise about what this is. The bill was released as a discussion draft, meaning it is an invitation for public comment and stakeholder feedback before formal introduction. That is a normal part of the legislative process, but it also means the final bill could look different from what was published.

Obernolte and Trahan have asked for input before the bill is formally introduced. Law firms, industry groups, and advocacy organisations are already submitting comments. The final text will be shaped by that feedback.

What This Means for Business

If you run a business that uses AI, the practical takeaway depends on how you use it.

If you deploy AI that makes decisions affecting employees, customers, or financial outcomes, the trajectory of federal regulation is pointing toward more disclosure and accountability. Even as a discussion draft, the Great American AI Act signals that Congress wants transparency around how AI models work and who audits them.

If you are thinking about using AI to reshape your workforce, the 60-day notification requirement for AI-attributable layoffs is worth flagging now, even before it becomes law. Building that kind of visibility into your AI deployment planning is good governance regardless of what the bill ultimately does.

And if you have been waiting for federal regulatory clarity before investing seriously in AI, the emergence of this draft suggests that clarity may be closer than it was a year ago. Congress is starting to take AI governance seriously. The question now is what the final framework looks like and how quickly it moves.

Want the practical version of this? The free Working With Claude field guide covers the full Claude ecosystem, Claude Code, and how to roll it out across a real business. Download it here.