The White House is negotiating a narrower state AI preemption package with Sen. Marsha Blackburn (R-Tenn.), bundling it with KOSA, the NO FAKES Act, and age-verification requirements, Axios reported June 8. The move effectively routes around the 269-page Great American AI Act discussion draft that Reps. Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.) released four days earlier.

The House draft is the more ambitious instrument. It proposes three years of preemption over state AI development laws and, per a document circulated by Trahan’s office, would knock out California’s AB 2013, parts of SB 942, and federalize frontier-safety regimes already on the books in California, New York, and Illinois. It’s also bleeding Democratic support over exactly that preemption clause, according to Axios. Deployment-side rules sit outside the carveout in both drafts, a detail the Future of Privacy Forum has been tracking closely.

Blackburn’s package is smaller, vaguer, and has the asset the last White House preemption push lacked: a Republican senator willing to put her name on it. Her spokesperson framed the deal as “not blanket preemption of all laws regulating AI or child safety.” Brad Carson of Americans for Responsible Innovation called preemption a “generational mistake.”

Running underneath all of this is Trump’s June 2 executive order standing up a voluntary AI cybersecurity clearinghouse led by Treasury with NSA and CISA, on a 30-day clock for OMB grant identification. The Christian Science Monitor framed it on June 10 as the opening move of national AI oversight, paired with OpenAI’s regulatory blueprint.

Obernolte’s office is still gathering stakeholder feedback. Blackburn’s package still has no legislative text. Industry trackers including LemonLime have noted the obvious dynamic: whichever bill lands first sets the preemption baseline, and Anthropic, OpenAI, and the states are all reading the same calendar.

Sources

Sources