This page lists every measure that would license, register, cap, or export-control language models, with its current status and our position. Each entry links to an official source and shows the date we last verified it.
Opposewe are working to defeat or amend itMonitorit touches our platform but we have not taken a positionSupportwe are working to pass itWondefeated, vetoed, or rescindedLostenacted over our objection
Colorado · SB 26-189 (replacing SB 24-205)
Oppose
Automated Decision-Making Technology Act (replacing the Colorado AI Act)
Repeals and re-enacts Colorado’s AI Act. Developers must supply deployers with documentation of intended uses, training-data categories, and limitations; deployers owe notice, explanations of adverse decisions, and human review. Drops the impact-assessment and duty-of-care provisions.
Our position. Compelled documentation of how a model was trained is compelled speech about a publication. The narrowing is welcome. The developer mandate remains.
Status. Signed May 14, 2026; effective Jan. 1, 2027. Attorney General rulemaking is underway, with the public hearing and final comment deadline on Oct. 26, 2026. Enforcement is stayed pending X.AI v. Weiser. (verified Oct 26, 2026)
Frontier AI provisions in the economic development bill
The Senate version requires developers above $500 million in revenue to maintain frontier frameworks, report incidents, submit to third-party review, and protect whistleblowers, with a commission reporting by March 1, 2027.
Our position. Third-party review as a condition of publishing is the Illinois audit model spreading. We urge conferees to strike the developer mandates.
Status. In conference committee since July 30, 2026. Not enacted. Status drawn from legislative press releases; confirm against the current conference report. (verified Jul 30, 2026)
Creates a federal regime for models trained above 10^26 operations: published frameworks, incident reporting, independent audits for the largest developers, and Commerce authority to intervene on a finding of “imminent catastrophic risk.” Narrowly preempts state frontier-transparency laws.
Our position. A compute threshold regulates speech by its length, and an intervention power over releases is a licensing scheme in waiting. Preemption of state rules is not worth reproducing them nationally.
Status. Introduced July 23, 2026 by Reps. Obernolte and Trahan; referred to Energy and Commerce and to Science, Space, and Technology. Sponsors sought a September markup; none has been confirmed. (verified Jul 23, 2026)
Artificial Intelligence Risk Evaluation Act of 2025
Establishes an Advanced AI Evaluation Program at the Department of Energy and conditions deployment of covered advanced systems on completing evaluation.
Our position. Conditioning deployment on a government evaluation is pre-publication review by another name. It is the clearest licensing proposal in Congress.
Status. Introduced Sept. 29, 2025 by Sens. Hawley and Blumenthal; referred to Senate Commerce. No cosponsors and no committee action. (verified Sep 29, 2025)
Advanced AI Accountability Act (“AEGIS Act”) draft
A circulated draft would make Commerce the frontier regulator, require third-party evaluations every 180 days and incident reports within 15 days, let a federal court suspend a model for 45 days, and preempt state law.
Our position. Court-ordered suspension of a model is a prior restraint on publication. We will oppose this text if it is introduced.
Status. Draft reported Oct. 16, 2026, last edited Sept. 29, 2026. No sponsor or bill number. (verified Oct 16, 2026)
Directs the Joint Commission on Technology and Science to study a framework for licensing independent verification organizations that check whether AI systems meet standards, and to report by Nov. 1, 2026. No developer mandates.
Our position. The report will shape Virginia’s 2027 session. We are watching for developer mandates in the recommendations.
Status. Signed Apr. 8, 2026. Report due Nov. 1, 2026. (verified Apr 8, 2026)
Executive Order 14365, Ensuring a National Policy Framework for Artificial Intelligence
Directs the Justice Department to form an AI Litigation Task Force against state AI laws, Commerce to list “onerous” state laws, and the administration to recommend preemption legislation. Carves out child safety, compute infrastructure, and state procurement.
Our position. We welcome federal challenges to state licensing regimes. We withhold support until the replacement framework is written, because preemption that re-regulates is no gain.
Status. Signed Dec. 11, 2025. The Task Force was created Jan. 9, 2026 and has intervened in one case, X.AI v. Weiser. The White House released a legislative framework Mar. 20, 2026; no preemption statute has been enacted. (verified Mar 20, 2026)
Prohibits AI developed or deployed with intent to manipulate behavior, discriminate, or produce unlawful deepfakes; adds government-use disclosures and a regulatory sandbox. Attorney-general enforcement with a cure period. No compute threshold or licensing.
Our position. Intent-based prohibitions on conduct, enforced against the actor, are the model we endorse. We watch how the attorney general applies it.
Status. Signed June 22, 2025; effective Jan. 1, 2026. (verified Jan 1, 2026)
Directs the Commerce Department to promote U.S. open-weight models and publish risk assessments of foreign-adversary open weights. Expressly grants no authority to ban a model.
Our position. It treats open weights as a national asset to promote rather than a hazard to license, and it writes the absence of ban authority into the text.
Status. Introduced Aug. 27, 2026 by Rep. Gabe Evans. Forwarded by subcommittee by voice vote Sept. 1, 2026 and reported by the full Energy and Commerce Committee, 26–21. (verified Sep 1, 2026)
Challenges Colorado’s AI Act as compelled speech under the First Amendment, and on dormant Commerce Clause, vagueness, and equal-protection grounds. The Justice Department intervened on the plaintiff’s side.
Our position. It is the first case to put a state AI developer mandate to a First Amendment test. We support the challenge and will track the merits briefing.
Status. Filed Apr. 9, 2026. DOJ intervened Apr. 24, 2026. Enforcement stayed by joint motion Apr. 27, 2026; the preliminary-injunction motion is due within 28 days of final Attorney General rulemaking. No merits ruling. (verified Apr 27, 2026)
American Artificial Intelligence Leadership and Uniformity Act
Imposes a temporary moratorium preempting state laws that restrict AI models in interstate commerce.
Our position. A patchwork of state licensing regimes is the most restrictive regime of all. A moratorium that preempts without re-regulating meets our sixth principle.
Status. Introduced Sept. 16, 2025 by Rep. Baumgartner; referred to the Energy and Commerce Subcommittee on Oversight and Investigations Dec. 19, 2025. No cosponsors and no further action. (verified Dec 19, 2025)
Commerce Department (BIS) · 90 Fed. Reg. 4544; ECCN 4E091
Won
Framework for Artificial Intelligence Diffusion
Created an export-control classification for the weights of closed models trained above 10^26 operations, with worldwide licensing tiers for chips and weights.
Our position. Controlling the export of a file of numbers is the encryption export regime of the 1990s repeated. Its rescission is a win. A replacement rule is still possible.
Status. Published Jan. 15, 2025. Rescinded by announcement May 13, 2025 before its compliance date. A redesigned rule went to regulatory review in February 2026 and was withdrawn in March 2026. No replacement has been published. (verified May 13, 2025)
Safe and Secure Innovation for Frontier AI Models Act
Would have required pre-training safety certifications, shutdown capability, third-party audits, and attorney-general enforcement for models above 10^26 operations or $100 million in training cost.
Our position. This bill is the template for every compute-threshold bill since. Its veto message criticized size-based triggers, which is our second principle in the governor’s words.
Status. Passed both houses; vetoed by Gov. Newsom Sept. 29, 2024. (verified Sep 29, 2024)
Models above 10^26 operations: incident reports within 72 hours, published frameworks, and, for developers above $500 million in revenue, the first mandatory annual third-party audits in the nation. Penalties of $1 million and $3 million.
Our position. A mandatory outside audit as a condition of continuing to publish is a license renewed every year.
Status. Signed July 6, 2026. Disclosure provisions effective Jan. 1, 2027; framework and audit provisions effective Jan. 1, 2028. (verified Jul 6, 2026)
New York · S6953-B / A6453-B; Ch. 96 of 2026 (amendments)
Lost
RAISE Act
Frontier developers above 10^26 operations must publish safety protocols, file transparency reports before deployment, report critical incidents to the Department of Financial Services within 72 hours, and, if revenue exceeds $500 million, produce quarterly catastrophic-risk assessments. Penalties of $1 million and $3 million.
Our position. This is the second state compute threshold, and it gives a financial regulator rulemaking power over publishing.
Status. Signed Dec. 19, 2025; chapter amendments signed Mar. 27, 2026. Effective Jan. 1, 2027. (verified Mar 27, 2026)
Transparency in Frontier Artificial Intelligence Act
Developers of models trained above 10^26 operations must report critical safety incidents within 15 days; those with revenue above $500 million must publish a frontier framework and transparency reports before deploying new models. Penalties up to $1 million per violation.
Our position. This is the first compute threshold enacted anywhere in the country. Mandatory pre-deployment reports are compelled speech tied to the size of a publication.
Status. Signed Sept. 29, 2025; effective Jan. 1, 2026. First annual state summary reports begin in 2027. (verified Jan 1, 2026)
Ten-year moratorium on state AI laws (budget reconciliation)
The House-passed reconciliation bill carried a ten-year moratorium on state AI regulation; the Senate version tied it to broadband funds.
Our position. We supported the moratorium as a bar on state licensing regimes. The Senate struck it.
Status. Struck by the Blackburn–Cantwell amendment, 99–1, on July 1, 2025. The bill was signed July 4, 2025 without it. A December 2025 attempt to attach preemption to the defense authorization bill also failed. (verified Jul 1, 2025)