Reps. Jay Obernolte (R-CA) and Lori Trahan (D-MA) released a 269-page discussion draft of the Great American AI Act on June 4, 2026, the most ambitious federal attempt yet to consolidate foundation-model oversight into a single regime. The price of consolidation: a three-year freeze on state laws regulating AI development.

The mechanism mirrors what California and a handful of other states have been assembling piecemeal. Frontier developers with prior-year revenue above $500 million must publish AI safety frameworks and submit to semi-annual audits conducted by state-licensed Independent Verification Organizations. Civil penalties reach $1 million per violation, per day, well above the $1M–$3M total caps in current state statutes (per IAPP). The bill also authorizes $100 million annually for a new Center for AI Standards and Innovation, codifies the National AI Research Resource at NSF, and extends the Cybersecurity Information Sharing Act of 2015 through fiscal 2035.

Cosponsors include Reps. Scott Franklin (R-FL), Suhas Subramanyam (D-VA), Erin Houchin (R-IN), and Scott Peters (D-CA). A Trahan aide told IAPP the introduced version will give IVO recommendations binding force, a meaningful strengthening from the draft.

The preemption clause is where the bipartisan veneer cracks. California’s AB 2013 training-data disclosure law and parts of the state’s RAISE Act-equivalent would be frozen out. Americans for Responsible Innovation is already running Massachusetts ads pressuring Trahan to drop the freeze. Brendan Steinhauser of the Alliance for Secure AI praised the bipartisanship but said federal action “does not justify preempting states’ ability to pass their own AI safeguards.” NetChoice’s Patrick Hedger called the federal standard “commendable.”

The split is structurally familiar. It’s the 1996 Telecommunications Act fight in miniature: a federal floor that industry calls coherence and state regulators call capture. Feedback goes to [email protected].

Sources

Sources