Congress demands AI companies reveal how their models work—and what could go wrong
H.R. 8094 — AI Foundation Model Transparency Act of 2026 · Filed by Donald Beyer (D-VA) · 3 cosponsors · Introduced Mar 26, 2026 · Referred to committee
Your members of Congress
Enter a ZIP to see where your representative and both senators stood on this bill.
Looked up on this device — your ZIP is never stored on our servers.
What it does
This bill requires the Federal Trade Commission to establish transparency rules for large AI foundation models—those with over 10 million users, trained on massive computing power, or capable of high-risk tasks. Companies must disclose training data sources, model limitations, safety testing results, and performance on sensitive benchmarks (medical, weapons, elections, hiring, etc.) either on their own websites or a central FTC registry. Small businesses get a 3-month grace period and compliance help. Open-source models are exempt.
Why we flagged it
The bill's core function is to impose disclosure and documentation requirements on large AI foundation model providers, enforced by the FTC as unfair/deceptive practice violations. It is fundamentally a transparency and consumer-protection regulation, not a subsidy, carve-out, or commemorative measure.
What the text implies
- The 10^26 computing-power threshold may inadvertently exempt smaller but still-capable models, creating a compliance cliff that incentivizes companies to stay just below the threshold rather than improve transparency across the board.
- Redaction provisions (cybersecurity, national security, federal law) are broad and undefined—companies may over-claim exemptions, limiting public visibility into actual model risks without clear FTC guidance on what qualifies.
The full analysis lists 5 implications of this text.
Who stands to gain
AI compliance and consulting firms; Standards-development organizations (NIST, IEEE); Cybersecurity and audit service providers