A 269-page federal AI bill landed in Congress on June 3, 2026 — and it has real teeth. The Great American AI Act, introduced by a bipartisan group of senators, would require companies deploying AI systems with over 1 million monthly active users to publish transparency reports, conduct third-party audits of high-risk AI systems, and establish liability frameworks for AI-caused harm. Here is what is inside, who gets hit hardest, and how it compares to the EU AI Act already in force across the Atlantic.
What the Bill Actually Requires
The core of the Great American AI Act runs across four sections. First, the Transparency Disclosure section requires companies to publish quarterly reports detailing what AI systems they deploy, what training data was used, and what safety evaluations were conducted. Second, the High-Risk AI Audit section mandates independent third-party audits for AI systems used in hiring, lending, healthcare, criminal justice, and critical infrastructure — the same categories the EU AI Act flags as highest risk. Third, the Liability Framework section establishes that AI developers and deployers share joint liability when a high-risk AI system causes "documented, quantifiable harm" — a significantly higher standard than existing tort law. Fourth, the Safe Harbor Provisions section exempts companies from certain liability claims if they can demonstrate compliance with an approved AI safety standard.
The bill defines "high-risk AI" as systems that make or assist decisions with legal, financial, health, or safety consequences for individuals — a definition broad enough to cover most enterprise AI deployments today. This is the kind of regulatory landscape we anticipated in our analysis of global AI regulation trends for 2026.
Who Gets Hit Hardest — And Who Gets Relief
OpenAI, Google, Microsoft, Meta, and Amazon are the bill's clear primary targets — each operates AI systems with hundreds of millions of monthly users. For these companies, the compliance calendar is tight: initial transparency reporting within 18 months of enactment, high-risk AI audit requirements within 24 months. Microsoft president Brad Smith has publicly said Microsoft "supports thoughtful AI regulation" without endorsing this specific bill. OpenAI has not publicly commented on the bill's specifics as of this writing. The bill includes explicit carve-outs for smaller AI companies — defined as those with under 100 employees or under $50 million in annual revenue — granting them a 36-month delayed compliance schedule and exemption from the third-party audit requirement until they exceed certain deployment thresholds.
How It Compares to the EU AI Act
The EU AI Act came into full force in 2026 with its risk-tier classification system — prohibited AI, high-risk AI, limited-risk AI, and minimal-risk AI. The Great American AI Act borrows this tiered structure but diverges in enforcement. The EU model is prescriptive: it specifies technical documentation requirements, conformity assessment procedures, and CE-marking equivalents. The US bill is more disclosure-centric and relies on the threat of private lawsuits under the new liability framework to drive compliance rather than a centralized EU-style enforcement agency. The bill proposes a new "AI Safety Office" within the FTC rather than a standalone agency — a more limited bureaucratic footprint than the EU's national authority network. For multinational companies, operating under both regimes simultaneously will be the central compliance challenge of 2027. As we covered in our piece on Colorado's SB 26-189 AI Act replacement, state-level AI laws are adding another layer of complexity on top of federal and EU requirements.
What Happens Next
The bill was referred to the Senate Commerce Committee on June 3, 2026. It has 14 co-sponsors — seven Democrats and seven Republicans — a bipartisan breadth unusual for major tech legislation. Committee hearings are expected in July 2026. The path to passage is not certain: the bill faces lobbying pressure from tech industry groups that support lighter-touch frameworks, and the House version has not yet been introduced. The most likely outcome before the end of 2026 is that a narrowed version focused specifically on transparency and high-risk AI audit requirements passes, with the liability and safe harbor provisions either removed or substantially modified.
Frequently Asked Questions (FAQs)
Q: What is the Great American AI Act and when was it released?
A: The Great American AI Act is a 269-page federal AI regulation bill introduced in Congress in June 2026. It establishes transparency requirements, liability rules, and mandatory auditing for high-risk AI systems used by large technology companies.
Q: Which companies are most affected by the Great American AI Act?
A: The bill primarily targets AI systems with more than 1 million monthly active users — OpenAI, Google, Microsoft, Meta, and Amazon are the primary targets. Smaller AI startups may qualify for lighter-touch compliance requirements.
Q: When would the Great American AI Act take effect?
A: The bill proposes a phased implementation. Large AI providers would have 18 months from enactment for initial compliance; high-risk AI system requirements kick in at 24 months. The bill has not yet passed.
Q: How does the Great American AI Act compare to the EU AI Act?
A: The EU AI Act uses a risk-tier classification system with prescriptive technical requirements. The Great American AI Act takes a similar tiered approach but focuses more on transparency disclosure and self-certification, with less prescriptive technical requirements than the EU version.
The Great American AI Act is the most consequential AI bill proposed in the United States to date. Whether it passes intact, gets narrowed, or stalls in committee will define the US AI regulatory landscape through 2027 and beyond. Watch this space — TechPopDaily will track every committee hearing and amendment as the bill moves forward.
