US AI Regulation in 2026: The Great American AI Act and What It Means for Businesses
A federal AI framework is finally taking shape
Representatives Jay Obernolte (R-CA) and Lori Trahan (D-MA) released a 269-page bipartisan discussion draft of the Great American Artificial Intelligence Act (GAAIA) on June 4, 2026 — the first comprehensive federal AI governance framework proposed in Congress. The bill would require frontier AI model developers to disclose model information and undergo third-party audits, and proposes a three-year preemption of state laws specifically regulating AI model development.
Congress is moving on multiple AI bills at once
On June 25, 2026, the House Science, Space, and Technology Committee marked up and favorably reported 10 AI-related bills in a single session with strong bipartisan support. This level of committee activity suggests federal AI legislation is progressing faster than the historically slow pace of US tech regulation.
The FTC is weighing in on AI accuracy
The FTC set a public comment deadline of July 31, 2026, for a policy statement addressing AI accuracy, specifically examining state laws that require alteration of the 'truthful outputs' of AI models — following a presidential executive order directing the FTC to issue guidance on the issue. Businesses building AI products with US-facing outputs should watch this closely, as it will shape how much liability attaches to AI-generated content.
States aren't waiting for Washington
At least 35 AI-related state bills have been enacted so far, including laws in Alabama and Georgia prohibiting health insurers from using AI to make adverse coverage decisions without sufficient human review. The proposed three-year state-law preemption in GAAIA — if passed — would significantly reshape this patchwork, so businesses should track both federal and state developments rather than assuming one will simply override the other soon.
The practical takeaway for businesses building AI products in the US
With federal, FTC, and state-level AI regulation all moving simultaneously, US businesses should build AI systems with documentation, human review checkpoints, and accuracy monitoring in place now — regardless of which specific bill ultimately passes — since these are common threads across nearly every proposal currently on the table.
Need help navigating what this means for your business?
Talk to our team about what's changing in AI and how it affects your roadmap — free consultation.
Get in Touch