AI Hiring Laws by State (2026): In Force, Enacted or Dead

Six states have enacted laws that govern how employers use artificial intelligence in hiring and other employment decisions: Illinois, Colorado, Connecticut, California, Texas and Maryland. New York City has the only law that requires a bias audit. Four of those rules are in force on October 19, 2026, Colorado's starts on January 1, 2027, and Connecticut's phases in between October 1, 2026 and October 1, 2027. Every other state relies on its existing discrimination statutes, and federal law has moved away from, not towards, AI-specific rules.

Counting is harder than it sounds, which is why the pages that rank for this question disagree with each other. Some count bills as laws (155 AI-in-the-workplace bills were introduced in 2026; Virginia's passed and was vetoed). Some count Utah's general AI disclosure law as a hiring law. Several still give Colorado's old date. This page states its counting rule, cites the statute for each row, and marks the figures we could not verify against the primary text.

The table

Counting rule: a row appears if a statute or regulation that has been enacted names artificial intelligence, automated decision tools or automated decision systems and places a duty on employers using them in employment decisions. Guidance documents, pending bills and general consumer AI laws are listed separately below.

Verification note: we read the enacted Colorado text and the New York City rule. The Illinois, California, Texas and Connecticut rows rest on the statute citations and law firm summaries because the legislature websites refused automated access during research. Penalty figures for Illinois and Connecticut are therefore described rather than quantified. A $50,000 Illinois penalty that circulates on some pages does not appear in the Act.

In force now

New York City

Local Law 144 is the only US rule that mandates an audit. An automated employment decision tool, defined in detail on our AEDT glossary page, may not be used to substantially assist or replace a hiring or promotion decision for a New York City candidate unless an independent auditor has calculated its selection or scoring rates by sex and by race and ethnicity within the preceding year, the employer has published a summary, and candidates were told 10 business days in advance. Enforcement began July 5, 2023.

Illinois

Illinois has two laws. The 2020 video interview act is narrow: it covers AI analysis of recorded interviews for Illinois positions and requires notice, an explanation of how the AI works, consent, and deletion within 30 days of a request. HB 3773 is broad. From January 1, 2026 it is a civil rights violation for an employer to use AI "with respect to recruitment, hiring, promotion, renewal of employment, selection for training or apprenticeship, discharge, discipline, tenure, or the terms, privileges, or conditions of employment" in a way that has a discriminatory effect on a protected class, to use zip codes as a proxy for protected class, or to fail to give notice of such use. The Department of Human Rights proposed notice rules in May 2026 and withdrew them in June 2026, so the statute's general notice duty applies without regulatory detail. Illinois' Biometric Information Privacy Act is a separate exposure for any tool that analyses voice or face.

California

The Civil Rights Council's regulations took effect October 1, 2025 and are the most consequential rules in force today, because they reach every employer with five or more employees that uses an automated decision system in any employment decision. They make it unlawful to use a system that discriminates, treat a vendor that designs or operates the system as the employer's agent, extend record retention to four years, and say that anti-bias testing, or its absence, is relevant to whether the employer has a defense. The CCPA's separate automated decision-making rules add notice, opt-out and risk assessment duties for CCPA-covered businesses from January 1, 2027. Our California AI workplace laws guide has the detail.

Texas

The Responsible Artificial Intelligence Governance Act, in force since January 1, 2026, is often listed as an AI hiring law and does little in hiring. It prohibits developing or deploying AI "with the intent to unlawfully discriminate", and states that disparate impact alone is not sufficient to show intent. Enforcement is by the Attorney General after a cure period. It adds no notice, audit or record duty.

Maryland

Since October 1, 2020, Maryland employers may not use facial recognition during an interview without the applicant's signed waiver. The waiver must state the applicant's name, the interview date, consent to the use of facial recognition, and that the applicant read the waiver. It is the narrowest law on this page.

Enacted, not yet in force

Colorado

The 2024 Colorado AI Act was repealed and re-enacted on May 14, 2026 as SB 26-189 after a federal lawsuit by x.AI, joined by the Department of Justice, paused enforcement. The new law drops impact assessments and the small-employer exemption and requires notice, a 30-day explanation after an adverse outcome, data correction, meaningful human review and three years of records, for decisions made on or after January 1, 2027. Our Colorado AI Act for employers guide quotes the enacted sections.

Connecticut

Connecticut's Public Act 26-15 was signed in spring 2026 (sources give dates between May 14 and June 2). From October 1, 2026, an employer cannot defend a discrimination claim by pointing to the use of an automated tool. From October 1, 2027, employers that deploy automated employment decision technology as a substantial factor in hiring, promotion, discipline or discharge must give notice and an explanation of adverse decisions. There is no bias audit mandate, though evidence of anti-bias testing is relevant. Enforcement is by the Attorney General as an unfair trade practice, with a 60-day cure period for violations through December 31, 2027.

Bills that passed and died, and the states with nothing

• Virginia. HB 2094, a Colorado-style high-risk AI bill, passed both chambers and was vetoed by Governor Youngkin on March 24, 2025. A 2026 successor had not been enacted at the time of writing.
• New York State. No statewide AI hiring law has been enacted. New York City's law covers only City candidates.
• New Jersey. The Attorney General and Division on Civil Rights have issued guidance that the Law Against Discrimination applies to algorithmic discrimination. Guidance is not a statute and adds no new duty.
• Utah. The Artificial Intelligence Policy Act requires disclosure that a person is interacting with generative AI when asked, and proactive disclosure in regulated occupations. It is a consumer law, not a hiring law, and sunsets July 1, 2027.
• Everywhere else. Title VII, the ADA, the ADEA and the state's own fair employment law apply to AI-assisted decisions. No notice, audit or record duty attaches.

What federal law says in 2026

There is no federal AI hiring statute. The Equal Employment Opportunity Commission's two technical assistance documents, on the ADA (May 2022) and on Title VII adverse impact (May 2023), were removed from its website in January 2025. Executive Order 14281 of April 23, 2025 directed federal agencies to deprioritise disparate impact enforcement. Neither action changes the statutes: Title VII's disparate impact provision, section 703(k), is still law and private plaintiffs can still bring claims under it, as the collective action in Mobley v. Workday in the Northern District of California shows. The Uniform Guidelines on Employee Selection Procedures, 29 CFR 1607.4(D), still describe a selection rate for any group below four-fifths of the highest group's rate as evidence of adverse impact. The four-fifths rule is a regulatory rule of thumb, not a safe harbor, and New York City's audit requirement uses the same ratio. A December 2025 executive order on pre-empting state AI laws does not itself override any state statute; Congress rejected a ten-year moratorium on state AI laws in 2025.

How enforcement has gone

Only New York City has an enforcement record to examine, and it is thin. The New York State Comptroller's audit of the Department of Consumer and Worker Protection, released December 2, 2025, found that the agency's enforcement was complaint-driven and its complaint routing "ineffective". Secondary reports of the audit say the agency reviewed 32 posted bias audits and found one issue where the Comptroller's staff found at least 17; we could not open the audit PDF to confirm those counts. A 2024 study by 155 student investigators, posted to arXiv, checked 391 employers and found 18 published audit reports, about 5%, and 13 transparency notices. SHRM's 2026 State of AI in HR research reported that 57% of HR professionals in states with AI workforce laws were unaware of them, a figure we have from SHRM's summaries rather than the gated report.

What a multistate employer does

1. Apply New York City's audit everywhere. If one covered tool is used on a New York City candidate, the audit must exist. Running it for the whole applicant pool costs little more and satisfies California's "relevant to a defense" standard.
2. Write one notice that meets Illinois, Colorado and Connecticut. Illinois requires notice of AI use in listed decisions now; Colorado requires a clear notice at the point of interaction from January 1, 2027; Connecticut requires notice from October 1, 2027. A single application-page notice naming the tool and its role covers all three.
3. Build the adverse-outcome letter. Colorado's 30-day letter and Connecticut's explanation of adverse decisions have the same content: what the decision was, what the tool did, how to request data and human review.
4. Keep records for four years. California's four years exceeds Colorado's three.
5. Designate a human reviewer with override authority. Colorado defines meaningful human review; it is good practice under every other regime and the best evidence in a Title VII case.
6. Write it into policy. Section 8 of our AI use policy template bars AI in employment decisions without legal sign-off and a named reviewer, and our AI resume screening guide covers the vendor questions.

Key takeaways

• Six states (Illinois, Colorado, Connecticut, California, Texas, Maryland) plus New York City have enacted laws placing duties on employers that use AI in employment decisions. Counts on other pages differ because they include bills, vetoed laws or general AI laws.
• New York City's Local Law 144 is the only rule requiring a bias audit: within one year before use, with a public summary and 10 business days' notice, enforced since July 5, 2023.
• Illinois HB 3773 (January 1, 2026) bars AI with a discriminatory effect across hiring, promotion, discipline and discharge, bans zip-code proxies and requires notice. The Department of Human Rights withdrew its proposed rules in June 2026.
• California's Civil Rights Council regulations (October 1, 2025) make vendors the employer's agent and require four years of records; CCPA automated decision rules follow on January 1, 2027.
• Colorado's re-enacted law (SB 26-189) applies to decisions from January 1, 2027 with notice, 30-day explanations and human review. Connecticut's Public Act 26-15 phases in on October 1, 2026 and October 1, 2027.
• Texas prohibits only intentional discrimination by AI; disparate impact alone is not a violation.
• Federal AI guidance from the EEOC was removed in January 2025. Title VII, the ADA, the ADEA and the four-fifths rule in 29 CFR 1607.4(D) are unchanged.
• A 2024 study of 391 New York City employers found about 5% had published the audit the law requires.