Every US state and DC, checked one at a time against the state's own legislature, code or agency — what each has actually enacted on artificial intelligence, what it covers in plain language, when it takes effect, and a link to the official source. Free. Informational, not legal advice.
| Jurisdictions with AI law already in force | 46 of 51 |
|---|---|
| Enacted but not yet in force | 1 — Georgia |
| No AI law found (verified, not assumed) | 3 — District of Columbia, Massachusetts, Missouri |
| Could not be verified | 1 — Alaska |
Last verified 29 September 2026. State AI regulation is no longer an early-adopter story — it is close to universal. What varies enormously is depth.
The bias-audit rule everyone cites — NYC Local Law 144 — is a New York City ordinance. New York State has no statewide AI hiring statute. This is the most repeated error in US AI-regulation coverage.
Virginia's general AI act was vetoed. Connecticut's died on the House calendar twice. North Dakota's deepfake bill failed 17–69. Virginia also has chapters that were signed but never took effect. Each was checked individually.
Colorado alone has a general algorithmic-discrimination regime, and it is now in force. Texas's act is intent-based with no private right of action; Utah's was narrowed to "disclose only if asked" and sunsets in 2027.
It's deepfakes and non-consensual imagery, then election synthetic media, then chatbot duties — the fastest-moving category, with a cluster landing 1 January 2027.
Every row links the state's own source. Where something could not be verified, the row says so rather than guessing — see the caveats below.
| Jurisdiction | Status | What it actually covers | Key law | Effective | Source |
|---|---|---|---|---|---|
| Alabama | In force | Law-enforcement AI limits (AI can't be the sole basis for an arrest); insurer AI in health-coverage decisions from Oct 2026 | SB 56 (2022) Act 2022-420; SB 63 (2026) Act 2026-589 | 2022-07-01 | official source → |
| Alaska not verified — see caveats | Could not verify | COULD NOT BE CHECKED — the state legislature's site blocks automated access. This is 'not verified', NOT 'no law'. | — | — | official source → |
| Arizona | In force | Election deepfakes: AI-disclosure on candidate deepfakes within 90 days of an election; digital-impersonation action (declaratory relief only) | A.R.S. §§16-1023, 16-1024 | 2024-05-21 | official source → |
| Arkansas | In force | Deepfake NCII with liability reaching the DEVELOPER of image-generation tools lacking safeguards; public-entity AI-use policies; generative-AI output ownership | Acts 827, 848, 927 (2025) | 2025 | official source → |
| California | In force | The broadest set: frontier-AI safety duties, training-data transparency, AI content provenance, plus CPPA automated-decision regulations | SB 53; AB 2013; SB 942/AB 853 | 2025-09-29 | official source → |
| Colorado | In force | THE ONLY general algorithmic-discrimination law in the US — duty of reasonable care for developers AND deployers of high-risk AI, impact assessments, consumer notice/appeal | SB24-205 (delayed by SB25B-004) | 2026-06-30 | official source → |
| Connecticut | In force | Government AI inventory + discrimination assessment; generative-AI subscription disclosure from Oct 2026. NOTE: its comprehensive AI bill died twice. | PA 23-16; PA 26-100 | 2023-07-01 | official source → |
| Delaware | In force | Government AI commission and inventory; election deepfake restrictions with a safe-harbour disclosure | HB 333; HS1 for HB 316 (2024) | 2024-07-17 | official source → |
| District of Columbia verified negative | No AI law found | NO AI LAW — verified by enumerating all 484 enacted DC laws across Council Periods 24-26. The Stop Discrimination by Algorithms Act has never passed. | — | — | official source → |
| Florida | In force | Political-ad AI disclaimer; takedown process for altered sexual depictions ('Brooke's Law') | Ch. 2024-126; Ch. 2025-133 | 2024-07-01 | official source → |
| Georgia | Enacted — not yet in force | Enacted but NOT yet in force: health-coverage decisions may not rest solely on AI (2027); conversational-AI disclosure and minor protections (2027) | SB 444 Act 411; SB 540 Act 518 | 2027-01-01 | official source → |
| Hawaii | In force | Deceptive AI election media; bans harmful AI imitations of voice/likeness; AI-companion disclosure and self-harm protocols | Acts 191/2024, 247/2026, 248/2026 | 2024-07-03 | official source → |
| Idaho | In force | Explicit synthetic media a crime (felony where a minor is depicted); conversational-AI safety duties arrive 2027 | Ch.105/2024, Ch.121/2026, S1297 | 2024-07-01 | official source → |
| Illinois | In force | EMPLOYMENT AI: discriminatory-effect AI in hiring is a civil-rights violation and zip codes are banned as a proxy; AI video-interview consent; no AI psychotherapy | HB 3773 (PA 103-0804); 820 ILCS 42; PA 104-0054 | 2026-01-01 | official source → |
| Indiana effective dates unverified | In force | State AI task force; 'fabricated media' in elections; AI-created images added to intimate-image offences | SEA 150; HEA 1133; HEA 1047 (2024) | 2024 (dates unverified) | official source → |
| Iowa | In force | Conversational-AI requirements (duties from 2027); health-carrier utilization review incl. AI; AI-modified depictions of minors | SF 2417; HF 2635; SF 2243 | 2026-07-01 | official source → |
| Kansas | In force | AI-generated/morphed CSAM; bans AI platforms controlled by a country of concern on state devices | HB 2183; HB 2313 | 2026-07-01 | official source → |
| Kentucky | In force | State-agency AI standards and an AI Governance Committee, plus a civil action for synthetic election media | SB 4 (2025 Acts Ch.66) | 2025-03-24 | official source → |
| Louisiana | In force | The most prolific deepfake state: AI nude-image coercion, election deepfake disclosure, AI-CSAM, and a ban on using a child's image to TRAIN AI | Acts 142/2024, 806/2026, 892/2026, 211/2026 | 2024-08-01 | official source → |
| Maine | In force | AI chatbots must not mislead a consumer into thinking they're human; no AI therapy unless by a licensed professional | LD 1727 (PL c.294); LD 2082 (PL c.687) | 2025-09-24 | official source → |
| Maryland | In force | One of the earliest EMPLOYMENT AI laws (no facial recognition in job interviews without waiver); election deepfakes; AI identity fraud from Oct 2026 | HB 1202 (2020) Ch.446; SB 141; SB 8 | 2020-10-01 | official source → |
| Massachusetts verified negative | No AI law found | NO AI LAW — full 2025 and 2026 Acts indexes checked. The 2024 Mass Leads Act mentions AI only as funding, not regulation. | — | — | official source → |
| Michigan | In force | Election-only: AI political-ad disclaimers, a statutory AI definition, and a misdemeanour for deceptive AI media near an election | PA 263/264/265 (2023) | 2024-02-13 | official source → |
| Minnesota | In force | Deepfake election crime, non-consensual deepfake crime, and a civil cause of action | HF 1370 (2023) Ch.58 | 2023-08-01 | official source → |
| Mississippi definition only, no duty | In force | DEFINITION ONLY — defines 'artificial intelligence' in state code and imposes no duty or restriction whatsoever | HB 1723 (2026) | 2026-07-01 | official source → |
| Missouri verified negative | No AI law found | NO AI LAW — roughly 20 AI bills in the 2026 session and not one passed. Volume of bills is not law. | — | — | official source → |
| Montana deregulatory | In force | DEREGULATORY: a 'Right to Compute' declaring computation a protected right, limiting state restriction. Plus government-AI limits and election deepfakes. | SB 212; HB 178; SB 25 (2025) | 2025-04-16 | official source → |
| Nebraska | In force | HEALTHCARE AI: an AI algorithm may not be the SOLE basis of a utilization-review denial, with disclosure and audit rights. Plus computer-generated CSAM and NCII. | LB 77 §12; LB 383; LB 371 (2025) | 2026-01-01 | official source → |
| Nevada | In force | Bars AI performing school counselling/psychology duties or posing as qualified for mental health care; emergency decisions must be made by a person | AB 406 Ch.283; AB 325; SB 263 (2025) | 2025-07-01 | official source → |
| New Hampshire | In force | Covers elections, government use and general deepfake harm simultaneously: candidate deepfake disclosure, state-agency AI limits, and a Class B felony for harmful deepfakes | Ch.345, Ch.209, Ch.243 (2024) | 2024-07-01 | official source → |
| New Jersey | In force | Deceptive audio/visual media is a third-degree crime, with a civil action; carve-outs for satire, news and AI developers | A3540, P.L.2025 c.40 | 2025-04-02 | official source → |
| New Mexico | In force | Election deepfake disclaimers, and an AI definition plus rulemaking mandate in the Nursing Practice Act | HB 182 (2024); HB 178 (2025) | 2024-05-15 | official source → |
| New York | In force | Frontier-AI oversight (RAISE Act), state-agency automated decisions need meaningful human review, and AI-companion self-harm protocols. NOTE: the famous NYC hiring-audit law is a CITY ordinance, not state law. | Ch.699/2025 RAISE; Ch.674/2024 LOADinG; Ch.58/2025 | 2026-03-19 | official source → |
| North Carolina | In force | Deepfake NCII/CSAM; K-12 AI policies with a public list of AI tools used in schools; an executive order on state AI governance | S.L.2024-37; S.L.2026-41; EO 24 | 2024-12-01 | official source → |
| North Dakota | In force | Its general AI law is five words: 'person' does not include AI — no legal personhood. Plus an election AI-disclosure label. NOTE: its deepfake bill FAILED 17-69. | HB 1361 (2023); HB 1167 (2025) | 2023-04-12 | official source → |
| Ohio | In force | K-12 school AI policies; from Oct 2026 Medicaid may not act on automated review without human confirmation | HB 96 (2025); SB 315 (2026) | 2025-09-30 | official source → |
| Oklahoma | In force | Crime of disseminating an artificially generated sexual depiction; insurance-department AI bulletin | HB 1364 (2025) c.23; OID Bulletin 2024-11 | 2025-11-01 | official source → |
| Oregon | In force | Election synthetic-media disclosure with civil penalties; deepfake NCII; companion-AI duties incl. a ban on misrepresenting training data, arriving 2027 | SB 1571; HB 2299; SB 1546 | 2024-03-27 | official source → |
| Pennsylvania | In force | Criminal 'digital forgery' of a real person's likeness; AI-generated CSAM; an executive order governing state generative-AI use | Act 35/2025; Act 125/2024; EO 2023-19 | 2025-09-05 | official source → |
| Rhode Island | In force | Four AI acts in one day: election deepfakes, no AI psychotherapy unless licensed, AI-companion self-harm protocols (2027), and AI-documentation notice to patients | P.L.2025 c.409; P.L.2026 c.373, c.375 | 2025-07-02 | official source → |
| South Carolina | In force | The thinnest coverage: digitally forged intimate images, and real-estate licensees responsible for AI-produced work | Act 37/2025; Act 204/2024 | 2025-05-12 | official source → |
| South Dakota | In force | AI-generated CSAM, election deepfakes with a prescribed AI label, and digitally fabricated intimate material as a felony | SB 79; SB 164; SB 41 | 2024-07-01 | official source → |
| Tennessee | In force | The ELVIS Act — the first US anti-voice-cloning law — plus political-ad deepfake disclaimers and a ban on AI posing as a mental health professional | Pub.Ch.588 (2024); Ch.625; Ch.647 (2026) | 2024-07-01 | official source → |
| Texas | In force | TRAIGA — a general AI act, but INTENT-based: disparate impact is expressly insufficient and there is no private right of action. Heavier duties fall on government bodies. | HB 149 TRAIGA; SB 1188; SB 1361 | 2026-01-01 | official source → |
| Utah | In force | An AI Policy Act creating an Office of AI Policy and a regulatory sandbox — but its disclosure duty was NARROWED in 2025 to 'only if asked', and the Act SUNSETS in July 2027 | SB 149 (2024); SB 226 (2025); HB 452 | 2024-05-01 | official source → |
| Vermont | In force | Government AI inventories; election deepfakes; the strictest mental-health AI rule found (no AI-delivered mental health services unless by a licensed professional); neurorights | Act 132/2022; Act 75, 156, 101 (2026) | 2022-07-01 | official source → |
| Virginia | In force | AI output may not be the sole basis for detention, prosecution, sentencing or parole. NOTE: its general AI act was VETOED, and its synthetic-media chapters were signed but NEVER took effect. | HB 1642 (2025) Ch.637 | 2025-07-01 | official source → |
| Washington | In force | Deepfake/NCII, forged digital likeness, campaign synthetic media and AI-CSAM now; from 2027 content provenance duties and chatbot rules banning named manipulative techniques | SB 5105; HB 1205; SB 5152; HB 1170; HB 2225 | 2023-07-23 | official source → |
| West Virginia | In force | A state AI task force (terminating 2027) and fabricated intimate images plus computer-generated CSAM | HB 5690 Ch.43; SB 198; HB 3187 Ch.38 | 2024-06-02 | official source → |
| Wisconsin source reached via proxy — see caveats | In force | Election AI disclosure with a per-violation forfeiture, and a virtual-CSAM felony (the statute is computer-generated in wording and never says 'artificial intelligence') | 2023 Act 123; 2023 Act 224 | 2024-03-23 | official source → |
| Wyoming | In force | A six-part omnibus: AI is no defence to an offence, conditional developer immunity, synthetic sexual material and AI-CSAM felonies, a ban on AI designed to promote self-harm, and limits on platforms using AI against political speech | HB 102 (2026) ch.91 | 2026-07-01 | official source → |
A tracker is only worth what its weakest entry is worth, so here are ours. Alaska could not be checked at all — its legislature site blocks automated access. That means not verified, not "no law". Indiana's effective dates are unverified: its site is a JavaScript-only application, and rather than assume Indiana's usual default we left them open. Wisconsin's pages were reachable only through a reader proxy — the URLs and content are official, but the retrieval was indirect, so treat those two entries as sourced-with-an-asterisk. Mississippi merely defines AI and imposes no duty, so counting it as regulation would overstate things. Wisconsin's virtual-CSAM statute never uses the words "artificial intelligence" — it is computer-generated-content law. And Montana runs the other way: its "Right to Compute" act is deregulatory, so any claim that states are uniformly restricting AI is false.
Insurance is deliberately out of scope here. Insurance-department AI bulletins are departmental guidance rather than statutes and don't appear on legislature sites; roughly nineteen of those department sites block automated access. Read insurance posture from our AI Insurance Watch tracker, not from silence on this page.
Two trackers, two jobs. AI Insurance Watch is the insurance-specific deep dive — NAIC model-bulletin adoption state by state, plus Colorado's binding insurance statute and regulations. This page is the full general picture: deepfakes, elections, employment, algorithmic discrimination, government use, healthcare, chatbots, provenance and general-purpose AI acts — everything the insurance tracker doesn't cover.
Leave your email and we'll tell you when this tracker records a new state law. No charge, and no other use of your address.
What this is. A tracked, source-cited record of publicly available state action on artificial intelligence, checked on 29 September 2026. What this is not. It is not legal advice, not a compliance determination, and not a law firm. It does not interpret statutes, does not tell you whether you are compliant, and is not a substitute for reading the linked official document or consulting counsel. If you spot an error, email carey@kcalbin.com — corrections are welcome and will be checked against the official source.