Record of Record The AI policy news desk
Today's Edition·Federal desk
Washington bureau

Federal

The White House, Congress, executive orders, and federal AI statutes — tracked from the Washington bureau.

Federal

Bipartisan House draft would freeze state AI-model laws for three years

The 269-page Great American AI Act discussion draft from Reps. Obernolte and Trahan pairs federal frontier-model rules with a preemption fight Congress just lost.

Why it matters: A three-year preemption of state AI-development laws would federalize California, New York, and Illinois frontier-model regimes — and reopen the fight that killed the Big Beautiful Bill's 10-year moratorium.
Read the brief →
Federal

Bipartisan House draft would preempt state AI development laws for three years

The Great American AI Act discussion draft, released June 4 by Reps. Obernolte and Trahan, proposes a federal frontier-AI framework — and a temporary ceiling on state model-development rules.

Why it matters: The 269-page draft turns the state-law floor on frontier AI into a federal ceiling for three years, the most consequential trade-off Congress has put on paper.
Read the brief →
Federal

Great American AI Act discussion draft lands, immediately splits Washington

The 269-page Obernolte-Trahan bill would create the first comprehensive federal AI regime — and its three-year state preemption is already drawing fire from both parties.

Why it matters: A three-year federal preemption of state AI development laws is on the table for the first time since the Senate stripped the identical provision from the Big Beautiful Bill 99–1 last year.
Read the brief →
Federal

Trump lifts Claude Mythos 5, Fable 5 export ban as California signs statewide Anthropic deal

Commerce ended an 18-day export-license requirement on Anthropic's most advanced models June 30–July 1, one day after Gov. Gavin Newsom made California the first state to deploy Claude government-wide at a 50% discount.

Why it matters: Washington and Sacramento moved in opposite directions on the same AI company in 48 hours — one lifting a security ban, the other cutting a landmark procurement deal — deepening the federal-state rift on AI governance.
Read the brief →
Federal

Trump lifts export ban on Anthropic's Mythos 5 and Fable 5

Commerce dropped an 18-day export control on Anthropic's frontier models June 30, capping a standoff that tested Trump's new voluntary frontier-AI framework.

Why it matters: The reversal establishes the first real precedent for how Trump's June 2 executive order will work in practice — voluntary early access, trusted-partner gating, and Commerce as the enforcement lever.
Read the brief →
Federal

Bipartisan House duo unveils 269-page Great American AI Act with three-year state preemption

Reps. Obernolte and Trahan released a discussion draft June 4 that would set federal frontier-model safety rules and freeze state AI development laws for three years.

Why it matters: The bill would convert the state-level frontier safety floor into a federal ceiling for three years — the most consequential preemption fight in AI policy this Congress.
Read the brief →
Federal

Great American AI Act drops as bipartisan discussion draft, with 3-year state preemption

Reps. Obernolte and Trahan released a 269-page bipartisan framework on June 4 that would create the first comprehensive federal AI regime and freeze state model-development laws for three years.

Why it matters: The draft is the most serious bipartisan swing yet at a federal AI law — and its three-year preemption of state AI development rules is already the fight that will define it.
Read the brief →
Federal

Bipartisan 'Great American AI Act' draft would freeze state AI laws for three years

Reps. Jay Obernolte and Lori Trahan released a 269-page discussion draft June 4 that would impose binding obligations on frontier labs and preempt state AI development laws.

Why it matters: A 3-year preemption tucked inside the first comprehensive federal AI bill revives the same fight that killed a 10-year state-law moratorium 99-1 last year — only now with bipartisan cover and binding rules on frontier labs.
Read the brief →
Federal

Bipartisan House draft floats federal AI framework with three-year state preemption

Reps. Obernolte and Trahan released a 269-page Great American AI Act discussion draft June 4 pairing frontier model safety rules with a sunsetted block on state AI development laws.

Why it matters: The narrower three-year preemption revives the federal-vs.-state fight Congress abandoned last summer — and critics say it converts the state-law floor into a federal ceiling.
Read the brief →
Federal

Trump's AI security order starts a 30-day clock — and leaves 'covered frontier model' undefined

A June 2 executive order asks frontier AI developers to voluntarily hand the government up to 30 days of pre-release access. Deadlines start ticking immediately.

Why it matters: The White House abandoned mandatory licensing but built the scaffolding — classified benchmarking, NSA-run designations, a Treasury clearinghouse — that could mature into binding oversight later.
Read the brief →
Federal

Bipartisan 'Great American AI Act' draft would freeze state AI laws for three years

Reps. Obernolte and Trahan's 269-page discussion draft sets up a federal AI framework — and a fight with state legislators over preemption.

Why it matters: A three-year federal ceiling on state AI development laws would override existing consumer, worker, and child-safety protections without a federal framework yet in force.
Read the brief →
Federal

Bipartisan House duo drops 269-page 'Great American AI Act' with 3-year state preemption

Reps. Jay Obernolte and Lori Trahan unveiled a discussion draft June 4 that would impose binding safety obligations on frontier AI labs and freeze state development laws for three years.

Why it matters: It's the first serious bipartisan vehicle for federal frontier AI rules — but the three-year preemption of state development laws is already drawing fire from safety advocates and 22 state AGs.
Read the brief →
Federal

Senate AI preemption package wobbles as states revolt, NO FAKES advances

Sen. Marsha Blackburn's bundle of KOSA, NO FAKES, and a three-year freeze on state AI laws is drawing fire from 200+ state legislators even as Judiciary moved the deepfakes bill June 18.

Why it matters: The White House's last-ditch vehicle to preempt state AI laws now hinges on a fragile Senate package — and a bipartisan revolt from statehouses is hardening before the text is even final.
Read the brief →
Federal

Sanders unveils $7 trillion AI sovereign wealth fund bill

The American AI Sovereign Wealth Fund Act would tax AI companies with $200M+ in sales, give the public a 50% stake, and pay every American a roughly $1,000 annual dividend.

Why it matters: A sitting senator just proposed the most aggressive AI ownership intervention in US history — half the equity of every major AI firm, transferred to a federally managed fund — and the idea has cross-spectrum sympathizers, including inside the Trump administration.
Read the brief →
Federal

Senate Judiciary unanimously advances NO FAKES Act to Senate floor

The bipartisan bill creating a federal IP right over voice and likeness cleared committee by voice vote June 18, with First Amendment concerns parked for floor debate.

Why it matters: A 15-cosponsor, voice-vote committee win makes NO FAKES the most viable piece of federal AI legislation in the 119th Congress — and the first to convert digital likeness into a true federal property right.
Read the brief →
Federal

Bipartisan House duo drops 269-page 'Great American AI Act' draft

Obernolte and Trahan unveiled the most comprehensive federal AI framework yet, with mandatory audits, whistleblower protections, and a three-year preemption of state AI development laws.

Why it matters: The draft turns the floor of state AI rules into a federal ceiling for three years — and puts frontier labs on a mandatory disclosure-and-audit track for the first time.
Read the brief →
Federal

Bipartisan House AI bill proposes three-year preemption of state laws

Reps. Jay Obernolte and Lori Trahan released a 269-page Great American AI Act discussion draft June 4, pairing frontier-model safety rules with a sunsetting block on state AI development laws.

Why it matters: It's the most comprehensive federal AI bill yet — and the preemption clause would override California's AB 2013 and parts of SB 53, setting up a fight with state AGs and consumer groups.
Read the brief →
Federal

Bipartisan House AI bill would freeze state laws for three years

The 269-page Great American AI Act draft from Reps. Obernolte and Trahan is the first comprehensive federal AI governance proposal — and its preemption clause is drawing fire.

Why it matters: A three-year freeze on state AI development laws would consolidate frontier-model oversight in Washington for the first time — and unwind safety statutes already on the books in California, New York, and Illinois.
Read the brief →
Federal

House drops 269-page 'Great American AI Act' with three-year state preemption

Obernolte and Trahan's bipartisan discussion draft would federalize transparency and audit rules from California, New York, and Illinois, and freeze state AI development laws for three years.

Why it matters: The draft converts the state-level floor on frontier AI rules into a federal ceiling — the first serious bid at a comprehensive U.S. AI framework, and the first to put state preemption on the table in legislative text.
Read the brief →
Federal

Bipartisan House draft would freeze state AI laws for three years

Reps. Obernolte and Trahan released a 269-page discussion draft June 4 that would impose binding federal obligations on frontier AI developers and preempt state development laws through 2029.

Why it matters: A federal floor would replace California's and New York's frontier-model rules with one regime, ending the patchwork that has shaped AI compliance for startups and small businesses adopting the technology.
Read the brief →
Federal

Bipartisan House draft would freeze state AI development laws for three years

The 269-page Great American AI Act, released June 4 by Reps. Obernolte and Trahan, pairs federal frontier-model audits with a three-year preemption of state AI development rules.

Why it matters: A bipartisan House draft would convert the patchwork of state AI development laws into a federal ceiling, federalizing California, New York, and Illinois frontier safety rules and freezing new state action through December 2029.
Read the brief →
Federal

Bipartisan 'Great American AI Act' draft would preempt state AI development laws for three years

Reps. Jay Obernolte and Lori Trahan released a 269-page discussion draft June 4 mandating frontier model audits, AI-layoff disclosures, and a narrow preemption of state AI development laws.

Why it matters: The draft is the first serious bipartisan attempt to set a binding federal frontier-AI regime — and it picks a fight with the states by preempting laws like California's AB 2013 and parts of SB 53.
Read the brief →
Federal

Bipartisan 'Great American AI Act' draft would freeze state AI laws for three years

Reps. Jay Obernolte (R-CA) and Lori Trahan (D-MA) released a 269-page discussion draft June 4 that mandates frontier-model audits and incident reporting — and preempts state AI development laws for three years.

Why it matters: It's the first serious bipartisan attempt at a comprehensive federal AI framework — and it would override the state laws California, New York, and Illinois have already passed.
Read the brief →
Federal

Bipartisan 'Great American AI Act' draft lands — with a three-year state preemption bomb

Reps. Jay Obernolte and Lori Trahan released a 269-page discussion draft June 4 that would impose federal frontier-model audits and freeze state AI development laws for three years.

Why it matters: The bill is the most ambitious bipartisan federal AI framework yet — but its three-year preemption clause would void California, New York, and Illinois frontier safety laws and has already split Democrats.
Read the brief →
Federal

Senate Judiciary sets June 18 markup on NO FAKES Act

The bipartisan bill would create the first federal property right over AI voice and likeness replicas, with $750,000-per-work platform liability.

Why it matters: A committee vote sends the strongest version yet of a federal deepfake property right toward the Senate floor — preempting state digital-replica laws and putting platforms on the hook for up to $750,000 per work.
Read the brief →
Federal

State AGs subpoena OpenAI as White House cuts preemption deal

A multistate subpoena landed June 12 — the same week Sen. Marsha Blackburn and the White House are negotiating to freeze state AI laws in exchange for KOSA, NO FAKES, and age verification.

Why it matters: State AGs are opening a sweeping AI probe at the exact moment Washington is trying to take that authority away from them.
Read the brief →
Federal

Bipartisan House draft would freeze state AI laws for three years

The 269-page 'Great American AI Act of 2026' from Reps. Obernolte and Trahan landed June 4 to near-universal opposition over its preemption clause.

Why it matters: The most ambitious bipartisan AI bill yet ties frontier-model audits to a three-year preemption of state development laws — the same trade Senate killed 99-1 last summer.
Read the brief →
Federal

Bipartisan House draft would preempt state AI laws for three years

Reps. Jay Obernolte and Lori Trahan released a 269-page Great American AI Act on June 4, mandating frontier-model audits while freezing state development rules — and immediately drawing fire from safety advocates and House Democrats.

Why it matters: Congress's most ambitious bipartisan AI bill ties mandatory frontier-lab oversight to a three-year ban on state development laws — reviving a preemption fight the Senate killed 99-1 last summer.
Read the brief →
Federal

White House and Blackburn cut around bipartisan AI bill with narrower preemption package

Sen. Marsha Blackburn is negotiating a state AI-law preemption deal bundled with KOSA, NO FAKES, and age verification — sidelining the House's Great American AI Act as preemption opposition mounts.

Why it matters: Two preemption vehicles are now live in Washington, and the one with White House backing is the narrower Senate package — not the 269-page House draft that's drawing fire from Democrats and safety groups.
Read the brief →
Federal

White House and Blackburn restart push to preempt state AI laws

Sen. Marsha Blackburn is negotiating a subject-matter preemption package with the Trump White House, sidelining the bipartisan House Great American AI Act.

Why it matters: Blackburn's involvement gives the White House a Senate vehicle it lacked last year — and signals the 269-page House draft is not the administration's preferred path before August recess.
Read the brief →
Federal

White House cuts deal with Blackburn to preempt state AI laws

The package pairs federal override of state AI rules with KOSA, the NO FAKES Act, and age verification — sidelining a bipartisan House framework.

Why it matters: Sen. Marsha Blackburn killed the last preemption push 99-1; her support now gives the White House the votes it has lacked for 18 months to override the growing state AI patchwork.
Read the brief →
Federal

Bipartisan House draft would freeze state AI laws for three years

Reps. Obernolte and Trahan unveiled a 269-page Great American AI Act discussion draft June 4, drawing immediate opposition from the House Democratic AI Commission and safety groups.

Why it matters: A bipartisan bill would convert the floor of state AI laws into a federal ceiling for three years — the most consequential preemption fight on AI to date.
Read the brief →
Federal

White House revives state AI preemption push, bundles it with KOSA and NO FAKES

Sen. Marsha Blackburn is leading talks to trade federal override of state AI laws for Trump backing of kids-safety and deepfake bills, sidelining the bipartisan Obernolte-Trahan draft.

Why it matters: A package deal — preemption for KOSA, NO FAKES, and age verification — gives the White House its first credible path to overriding state AI laws after last year's failed attempts.
Read the brief →
Federal

Bipartisan House draft would freeze state AI development laws for three years

Reps. Obernolte and Trahan unveiled a 269-page Great American AI Act draft on June 4 requiring semi-annual audits of frontier labs and codifying CAISI at $100M a year.

Why it matters: It's the first comprehensive bipartisan House framework on AI — and it would nationalize the state-level model while wiping out California and Illinois development laws for three years.
Read the brief →
Federal

Trump signs AI executive order with 30-day voluntary review of frontier models

The June 2 order directs NSA to build classified benchmarks for 'covered frontier models,' stands up an AI cybersecurity clearinghouse at Treasury, and asks labs to hand over their most powerful models 30 days before release.

Why it matters: After scrapping a May version over innovation concerns, the White House just produced its first direct engagement with pre-deployment evaluation of frontier AI — and handed the evaluation keys to the NSA.
Read the brief →
Federal

Bipartisan 'Great American AI Act' would freeze state AI development laws for three years

Reps. Jay Obernolte and Lori Trahan released a 269-page discussion draft June 4 that preempts state AI development laws, codifies a federal standards body, and immediately drew fire from safety groups.

Why it matters: A three-year federal ceiling on state AI development laws would freeze California, New York, and Illinois frontier safety regimes in place — and the backlash from safety groups and a House AI commission landed before the ink dried.
Read the brief →
Federal

House drops 269-page 'Great American AI Act' — with a 3-year freeze on state laws

Reps. Obernolte and Trahan unveiled the first comprehensive federal AI governance discussion draft June 4, preempting state AI development laws for three years and drawing immediate fire from a House Democratic commission.

Why it matters: This is Republicans' most viable federal AI vehicle before the midterms, and it ties mandatory frontier safety rules to the third congressional attempt to override state AI laws.
Read the brief →
Federal

Bipartisan House duo drops 269-page 'Great American AI Act' with 3-year state preemption

Reps. Jay Obernolte (R-CA) and Lori Trahan (D-MA) unveiled a June 4 discussion draft creating the first comprehensive federal AI framework — and freezing state AI development laws for three years.

Why it matters: The bill would nationalize the foundation-model regulatory model being built state-by-state, replacing it with semi-annual third-party audits and $1M-per-day penalties — but only by killing state development laws for three years.
Read the brief →
Federal

Trump signs AI cyber review order after May collapse, with 30-day window

The 'Promoting Advanced Artificial Intelligence Innovation and Security' order, signed June 2 in a private ceremony, gives frontier labs a voluntary track to hand models to NSA, CISA, and Treasury for classified benchmarking.

Why it matters: After scrapping a 90-day version in May, the White House landed a narrower voluntary framework that explicitly forbids licensing or preclearance — a deliberate ceiling on how far federal AI oversight can go without Congress.
Read the brief →
Federal

Trump pulls AI cybersecurity executive order hours before signing

A voluntary pre-release review framework for frontier AI models was shelved May 21 after pushback from David Sacks, tech executives, and the president himself.

Why it matters: The accelerationists won this round — a White House push to put frontier models through voluntary federal security review collapsed at the signing table, leaving no executive framework for evaluating cyber-capable AI.
Read the brief →
Federal

Trump postpones AI executive order, says it 'gets in the way'

Order would have created a voluntary pre-release review process for advanced AI models, with federal agencies given up to 90 days for security review.

Why it matters: The postponement leaves the federal pre-release review question — central to every frontier lab's planning since the Mythos disclosure — unresolved, with no successor timeline.
Read the brief →