Bipartisan AI Bill Aims to Preempt State Regulations

On June 4, 2026, Reps. Jay Obernolte (R-CA) and Lori Trahan (D-MA) released a 269-page draft creating the Center for AI Standards and Innovation (CAISI) at NIST, requiring OpenAI, Anthropic and Google DeepMind to file safety plans, undergo audits and report incidents. The bill would preempt state laws on AI model development for three years, after which states could regulate again.

Politico

President Trump’s June executive order calls for voluntary 30-day model reviews, while state AI statutes exist in California, New York and Illinois. With the 2026 midterms approaching, lawmakers argue a federal floor can prevent regulatory patchwork.

Politico

Co-sponsors include Reps. Suhas Subramanyam (D-VA), Scott Peters (D-CA), Scott Franklin (R-FL) and Erin Houchin (R-IN). CAISI will oversee compliance; the Cybersecurity and Infrastructure Security Agency will aid code security. Advocacy groups include Americans for Responsible Innovation, Alliance for Secure AI, NetChoice and Tech Oversight Project.

Politico

The draft allocates $300 million to CAISI for three years and includes a three-year sunset on the preemption clause, after which states may again regulate AI model development.

Politico

Obernolte said the draft is released to gather stakeholder input before formal filing. NetChoice’s Patrick Hedger called the framework “strong” and praised its federal standard, urging clearer audit rules. House Speaker Mike Johnson supported preempting state AI rules to protect children. The administration’s executive order seeks voluntary model reviews, and Sen. Rand Paul has blocked renewal of the cyber-threat data-exchange provision.

Politico

Brad Carson of Americans for Responsible Innovation warned that eliminating state authority would be a “generational mistake.” Brendan Steinhauser of Alliance for Secure AI said the preemption “falls short,” urging a federal floor that still allows stronger state safeguards. Massachusetts and New York legislators have cautioned Trahan, and Democratic AI caucus says the draft does not meet the “enormity of the moment.” Florida Gov. Ron DeSantis called the plan an “AI amnesty.”

Politico

The draft’s language is ambiguous: some analysts say it bars any state regulation of AI model development, while the text also states states may continue to regulate AI use and deployment, leaving the exact scope unsettled.

Politico

• “We are releasing this draft to hear from stakeholders, experts, and the public so we can strengthen the legislation before it is formally introduced,” — Jay Obernolte, Representative (R-CA)

Reuters

• “I am unwilling to sit here and allow something as transformative as AI wreak havoc without any guardrails on my watch,” — Lori Trahan, Representative (D-MA)

Boston Globe

The draft remains open for public comment before formal House filing after the August recess.

Reuters