There is no comprehensive federal AI statute. Federal policy runs through executive orders and existing agency authority. The current administration's stance is deregulatory and centralizing: it revoked the prior administration's AI-safety order, issued an AI Action Plan in July 2025, and in December 2025 signed Executive Order 14365, which seeks to discourage and challenge state AI laws and to build toward a single federal framework that would preempt them. Agencies rely on the FTC Act, sector rules, and the NIST AI Risk Management Framework rather than a dedicated law.
Key points
- No comprehensive federal AI statute exists as of October 2026.
- The TAKE IT DOWN Act (May 19, 2025) criminalizes publishing non-consensual intimate images, AI-generated ones included, and has required platforms to remove reported images within 48 hours since May 19, 2026.
- Executive Order 14148 (January 20, 2025) revoked the prior administration's AI-safety order, EO 14110, and Executive Order 14179 (January 23, 2025) directed agencies to remove barriers to AI.
- America's AI Action Plan (July 23, 2025) set a deregulatory, pro-adoption agenda.
- Executive Order 14365 (Dec 11, 2025) created a DOJ AI Litigation Task Force and a strategy to preempt or defund 'onerous' state AI laws.
- State laws remain in force unless and until a court strikes them down or Congress preempts them.
- The Justice Department's AI Litigation Task Force made its first move in April 2026 by intervening in xAI's suit against the Colorado AI Act. Colorado agreed to a court order holding off enforcement, which also covers any 2026 replacement law, and then repealed and replaced the act. The task force has not won a ruling against any state AI law. Separately, on September 16, 2026, a federal judge barred Montana from enforcing its election deepfake law against a former legislator and his political committee, in a private First Amendment suit.
- No comprehensive federal AI bill has passed either chamber. The FRONTIER Act (H.R. 9925), introduced July 23, 2026 by Reps. Obernolte and Trahan, would set tiered transparency, audit, and incident-reporting duties for frontier developers and preempt state rules on the same subjects, and a Senate Commerce draft from Sens. Thune and Klobuchar stalled over whether companies may test their own models. No floor votes are expected before the midterm elections.
- Executive Order 14434 (September 29, 2026) tells federal agencies to use the terms 'super intelligence' and 'SI' in place of 'artificial intelligence' and 'AI' in non-statutory documents and asks the science adviser for proposed legislative language on a federal definition within 60 days. It creates no duties for companies; California answered the next day with its own order keeping the term AI. The same day, six frontier companies signed a voluntary White House accord on internal controls, independent audits, and oversight committees for their models, and on October 4 the President announced a Super Intelligence Force, a coordinating body with no published order, chaired by the Director of National Intelligence.
- On September 24, 2026, 25 state attorneys general from both parties wrote to the four congressional leaders urging Congress to regulate AI while preserving state oversight.
- Treasury's proposed rules under the COINS Act, announced October 22 and published in the Federal Register on October 26, 2026, would bar US investment in the development of any super intelligence (SI) system by a covered foreign person in a country of concern, using Executive Order 14434's term for AI; comments are due December 7, 2026.
State law is where the action is
With Congress silent, the states are writing the rules. Eleven are on the tracker, and they do not agree with each other. They include broad acts, single-issue statutes, a statutory right to compute, and one veto. The full comparison, sortable and filterable, lives in the state tracker.
Open the US state tracker →
Laws and instruments
Executive Order 14179, Removing Barriers to American Leadership in AI
EO 14179 (Jan 23, 2025)
Directed agencies to remove regulatory barriers to AI development and adoption and to review actions taken under the 2023 AI executive order, which Executive Order 14148 revoked on January 20, 2025.
Effective: Jan 23, 2025
TAKE IT DOWN Act
Pub. L. 119-12 (S. 146), signed May 19, 2025
Makes it a federal crime to knowingly publish non-consensual intimate images, including AI-generated 'digital forgeries', and requires platforms that host user content to remove a reported image within 48 hours and make reasonable efforts to remove known copies. The FTC enforces the platform duties as unfair or deceptive practices.
Effective: May 19, 2025; platform notice-and-removal duties from May 19, 2026
America's AI Action Plan
Released July 23, 2025
A whole-of-government plan directing agencies to remove barriers to AI, expand federal AI use, and treat AI leadership as a national priority.
Effective: July 23, 2025
Executive Order 14365, Ensuring a National Policy Framework for AI
EO 14365, 90 Fed. Reg. 58499 (Dec 11, 2025)
Directs the Attorney General to run an AI Litigation Task Force challenging state AI laws, tells Commerce to identify 'onerous' state laws, conditions some federal funding on state compliance, and asks the FTC and FCC to act. The legislative recommendation it calls for may not propose preempting state laws on child safety, AI compute and data-center infrastructure (other than generally applicable permitting reforms), or state government procurement and use of AI.
Effective: Dec 11, 2025
NIST AI Risk Management Framework
NIST AI 100-1; Generative AI Profile NIST-AI-600-1
The de facto US technical baseline. Voluntary, but referenced by state laws (Texas TRAIGA gives a defense to companies that substantially comply with it and find violations through internal review) and by federal contracting expectations.
Effective: Voluntary; widely referenced
Executive Order 14434, Inaugurating the Era of Super Intelligence
EO 14434 (September 29, 2026), 91 Fed. Reg. 63129
Makes it executive-branch policy to use the terms 'super intelligence' and 'SI' instead of 'artificial intelligence' and 'AI' in official correspondence, public communications, websites, reports, and other non-statutory documents, without altering existing regulations, contracts, or statutes. Defines SI by reference to the statutory definition of artificial intelligence in 15 U.S.C. 9401(3) and directs the Assistant to the President for Science and Technology to submit, within 60 days, proposed legislative language for a federal definition and whether it should supersede the statutory one. It imposes no obligations on companies. It was signed the same day as a voluntary accord in which Google, Anthropic, Meta, OpenAI, xAI, and NVIDIA committed to internal capability and alignment controls, an internal oversight team, independent external auditors, and an independent committee receiving their reports; NIST has since renamed its AI standards center the Center for Advancing Innovation and Standards for Super Intelligence.
Effective: September 29, 2026; legislative language due by November 28, 2026
Outbound investment rules under the COINS Act (proposed)
Treasury proposed rule, 91 Fed. Reg. 67396 (October 26, 2026); 31 CFR Parts 850 and 851
Treasury's proposed rules implementing the Comprehensive Outbound Investment National Security (COINS) Act would list the development of any super intelligence (SI) system by a covered foreign person in a country of concern, such as China, as a prohibited technology for US investors, with no technical threshold and no notification tier. The proposal uses the label SI system because Executive Order 14434 tells agencies to say SI in place of AI, and defines it much as the existing rules define an AI system, adding systems that generate content and covering agent software, robots, and other embodied systems that run on them. Treasury's existing outbound rules, in force since January 2, 2025, sort AI investments into prohibited and notifiable tiers by training compute and end use.
Effective: Not in force; comments due December 7, 2026
NO FAKES Act of 2026 (S. 4591)
S. 4591, ordered reported by the Senate Judiciary Committee June 18, 2026 and placed on the Senate calendar June 24, 2026 (Calendar No. 446); House companion pending
A federal intellectual-property right in every person's voice and visual likeness, with civil liability for producing or distributing unauthorized digital replicas and for platforms that host them with knowledge, at up to $750,000 per embodiment for platforms. Advanced unanimously by voice vote and awaiting Senate floor time. It would preempt future state digital-replica laws while preserving statutes and common-law claims in existence as of January 2, 2025, so Tennessee's ELVIS Act survives. Reported to be part of a broader preemption package under negotiation with the White House alongside the Kids Online Safety Act.
Effective: Not in force; committee-reported, Senate floor pending
Timeline
- January 20, 2025
EO 14148 revokes the prior AI-safety order, EO 14110.
- January 23, 2025
EO 14179 directs agencies to remove barriers to AI.
- May 19, 2025
TAKE IT DOWN Act signed: a federal crime to publish non-consensual intimate images, AI-generated ones included.
- July 23, 2025
America's AI Action Plan released.
- December 11, 2025
EO 14365 launches the state-preemption strategy.
- January 9, 2026
The Attorney General's memo establishes the DOJ AI Litigation Task Force.
- March 11, 2026
Commerce evaluation of state laws and FTC policy statement due.
- April 24, 2026
DOJ intervenes in xAI's suit against the Colorado AI Act, the task force's first action against a state law.
- May 19, 2026
TAKE IT DOWN Act platform duties apply: removal within 48 hours of a valid request. The FTC begins enforcing them and opens TakeItDown.ftc.gov for complaints; the week before, its chairman wrote to 15 major platforms reminding them of the deadline.
- June 2, 2026
Executive Order 14409 signed: a voluntary 30-day pre-release vetting framework for covered frontier models, an AI cybersecurity clearinghouse, and criminal-enforcement priority for AI-enabled attacks.
- June 18, 2026
NO FAKES Act (S. 4591) advanced unanimously by the Senate Judiciary Committee; a federal likeness right with preemption of future state digital-replica laws.
- September 15, 2026
No comprehensive federal AI bill has passed either chamber; Senate talks stall over frontier-model safety testing.
- September 17, 2026
An FTC consumer alert directs people to TakeItDown.ftc.gov to report platforms that miss the 48-hour removal deadline, citing civil penalties of up to $53,088 per violation.
- September 24, 2026
Twenty-five state attorneys general write to congressional leaders urging federal AI rules that preserve state oversight.
- September 29, 2026
Executive Order 14434 directs federal agencies to say 'super intelligence' instead of 'artificial intelligence'; published October 2.
- September 29, 2026
Six frontier companies sign a voluntary White House accord on model safety controls, independent audits, and oversight committees.
- October 4, 2026
The President announces a Super Intelligence Force to coordinate federal AI efforts; no executive order or memorandum has been published.
- October 20, 2026
DHS publishes in the Federal Register its Federal Acquisition Supply Chain Security Act order of July 24, 2026, which excludes DeepSeek products from federal civilian procurement and orders their removal, along with third-party products developed using DeepSeek, from civilian agency systems.
- October 26, 2026
Treasury's proposed COINS Act rules, announced October 22, are published; they would prohibit US investment in the development of any SI system in a country of concern, with comments due December 7.
Who it applies to
Federal action targets agencies and the federal-state relationship, and no comprehensive statute governs private conduct. The TAKE IT DOWN Act reaches platforms that host user content. Companies are governed mainly by state law plus generally applicable federal rules (FTC deception authority, sector regulators, civil-rights law).
Penalties
No AI-specific federal penalties. Exposure runs through the FTC Act, sector regulators, and civil-rights and consumer-protection statutes. 'AI washing' (overstating what a system does) is an active FTC enforcement theme. Under the TAKE IT DOWN Act, a platform that fails to remove a reported intimate image within 48 hours commits an unfair or deceptive practice; the FTC has taken complaints at TakeItDown.ftc.gov since enforcement began on May 19, 2026 and puts the civil penalty at up to $53,088 per violation. Publishing non-consensual intimate deepfakes carries up to two years in prison, three if the victim is a minor.
Recent developments
October 26, 2026
Federal Register
Treasury proposes to bar US investment in 'super intelligence' development in countries of concern
Treasury's proposed rules under the Comprehensive Outbound Investment National Security Act, announced October 22 and published October 26, would make the development of any super intelligence (SI) system by a covered foreign person in a country of concern, such as China, a prohibited technology for US investors. The proposal moves AI out of the notification tier of the existing outbound rules, sets no technical threshold, uses the term the September 29 executive order prescribes for AI, and adds biotechnology to the covered sectors. Comments are due December 7, 2026.
October 4, 2026
Associated Press
A 'Super Intelligence Force' and a voluntary industry accord follow the renaming order
The President announced on social media a Super Intelligence Force to coordinate the federal government's AI effort and reach out to AI companies, consumers, public interest and religious groups, and critical-infrastructure providers, reporting to him and the chief of staff. The Director of National Intelligence leads it, with the FTC chairman, the Pentagon's chief technology officer, and the OPM director as members. No executive order or memorandum has been published. It follows the September 29 White House meeting at which Google, Anthropic, Meta, OpenAI, xAI, and NVIDIA signed a voluntary accord committing to internal capability and alignment controls, an internal oversight team, independent external auditors, and an independent committee receiving their reports, with no enforcement mechanism. NIST has renamed its AI standards center the Center for Advancing Innovation and Standards for Super Intelligence.
September 29, 2026
Federal Register
Executive Order 14434 renames AI 'super intelligence' across the federal executive branch
The order makes it administration policy to use 'super intelligence' and 'SI' in place of 'artificial intelligence' and 'AI' in official communications and non-statutory documents, defines the term by reference to the existing statutory definition of AI, and gives the science adviser 60 days to propose legislative language for a federal definition. It changes no regulation, contract, or statute and imposes nothing on companies. California's governor responded on September 30 with Executive Order N-10-26 directing state agencies to keep using the term AI.
September 17, 2026
Federal Trade Commission
The FTC points victims to its TAKE IT DOWN Act complaint portal and names the penalty
A consumer alert told people to report a platform that fails to remove an intimate image within 48 hours of a valid request at TakeItDown.ftc.gov, open since the platform duties took effect on May 19. The agency says each violation can draw a civil penalty of up to $53,088. No enforcement action has been announced.
September 15, 2026
Roll Call
Congress remains far from an AI law as Senate talks stall over safety testing
No comprehensive federal AI bill has passed either chamber. A House bill from Reps. Obernolte and Trahan would require frontier developers to publish catastrophic-risk frameworks and would preempt state transparency, incident-reporting, and audit rules. In the Senate, a Commerce Committee draft from Sens. Thune and Klobuchar, shaped by Chair Cruz, lets companies test their own models and report results to the Commerce Secretary, while Sen. Cantwell wants testing at national labs and opposes preemption. No floor votes are expected before the midterm elections.
April 24, 2026
Jenner & Block
The AI Litigation Task Force's first move: joining xAI against Colorado
About three and a half months after it began operating, the Justice Department's task force intervened in xAI's federal suit against the Colorado AI Act, arguing on equal-protection grounds. Colorado agreed to a court order holding off enforcement, which also covers any 2026 replacement law, and then repealed and replaced the act in May.
March 11, 2026
Ropes & Gray
US federal preemption push meets legal limits as agency deadlines arrive
Analysis of Executive Order 14365 concludes the order cannot preempt state AI law on its own, since preemption generally flows from Congress. The FTC policy statement and Commerce evaluation of state laws were due this week, but courts will decide whether any of it actually displaces state rules.
December 11, 2025
The White House
Trump signs Executive Order 14365 to challenge and preempt state AI laws
The order creates a DOJ AI Litigation Task Force to challenge state AI laws, directs Commerce to flag 'onerous' laws, and conditions some federal funding on state compliance. It names Colorado's AI Act as an example and carves out state laws on child safety, data centers, and government procurement.
Common questions
Does the United States have a comprehensive federal AI law?
No. There is no comprehensive federal AI statute. Federal policy runs through executive orders and existing agency authority. The current administration's stance is deregulatory and centralizing: it revoked the prior administration's AI-safety order, issued an AI Action Plan in July 2025, and in December 2025 signed Executive Order 14365, which seeks to discourage and challenge state AI laws and to build toward a single federal framework that would preempt them. Agencies rely on the FTC Act, sector rules, and the NIST AI Risk Management Framework rather than a dedicated law.
What are the penalties for AI violations under US federal law?
No AI-specific federal penalties. Exposure runs through the FTC Act, sector regulators, and civil-rights and consumer-protection statutes. 'AI washing' (overstating what a system does) is an active FTC enforcement theme. Under the TAKE IT DOWN Act, a platform that fails to remove a reported intimate image within 48 hours commits an unfair or deceptive practice; the FTC has taken complaints at TakeItDown.ftc.gov since enforcement began on May 19, 2026 and puts the civil penalty at up to $53,088 per violation. Publishing non-consensual intimate deepfakes carries up to two years in prison, three if the victim is a minor.
Who does US federal AI regulation apply to?
Federal action targets agencies and the federal-state relationship, and no comprehensive statute governs private conduct. The TAKE IT DOWN Act reaches platforms that host user content. Companies are governed mainly by state law plus generally applicable federal rules (FTC deception authority, sector regulators, civil-rights law).
Cite this page
Artificial Intelligence Regulations, "AI Regulation in the United States," reviewed October 26, 2026, https://artificialintelligenceregulations.com/jurisdictions/united-states-federal.html.
Entries state the position as of the review date and link to the primary text for verification.
