AI in Hiring Laws by State and City (2026)

Only four states have a law or rule written for AI in employment decisions. The 44 in the last column (43 states plus DC) have no specific rule we found. That is not the same as no risk: ordinary discrimination law covers biased software too.

Use climbs with headcount, from 2.9% of the smallest firms to 14.8% of businesses with 100-249 employees, then dips slightly to 13.6% at 250 or more. The bigger the hiring volume, the more likely a screening tool is part of it.

Planned use runs far ahead of current use. Among businesses with 100-249 employees that plan to use AI, 47.4% expect to apply it to HR. Plenty of employers will be reading their first hiring software contract soon.

States with rules on the books sit close to the 4.2% national figure, from 3.3% in New Jersey to 5.0% in Colorado. These numbers can't show cause and effect, but they give no sign that the rules keep employers away. What changes is the paperwork that comes with the tool.

In Pew's 2022 survey, 71% opposed AI making the final hiring call and 66% said they would not apply to an employer that used it for hiring decisions. That is a candidate-relations issue as much as a legal one, and it is why notice rules matter.

Main AI hiring laws and rules in the U.S. (as of October 2026)

Source: New York City Council, Local Law 144 of 2021 (automated employment decision tools), legislation detail and text; NYC Department of Consumer and Worker Protection, Automated Employment Decision Tools (AEDT); NYC Department of Consumer and Worker Protection, Automated Employment Decision Tools: Frequently Asked Questions; California Civil Rights Department, Civil Rights Council Secures Approval for Regulations to Protect Against Employment Discrimination Related to Artificial Intelligence; California Privacy Protection Agency, CCPA Updates, Cybersecurity Audits, Risk Assessments, Automated Decisionmaking Technology (ADMT), and Insurance Companies regulations; California Privacy Protection Agency, Text of Regulations: CCPA Updates, Cybersecurity Audits, Risk Assessments, ADMT and Insurance Companies (11 CCR § 7200); New Jersey Office of the Attorney General, Division on Civil Rights, N.J.A.C. 13:16, Disparate Impact Discrimination (R.2025 d.150); Texas Legislature, HB 149 enrolled text, Texas Responsible Artificial Intelligence Governance Act; Texas Legislature Online, HB 149 bill history; Maryland General Assembly, Maryland Code, Labor and Employment § 3-717 (facial recognition services in interviews); Colorado General Assembly, SB24-205 Consumer Protections for Artificial Intelligence; Colorado General Assembly, SB25B-004 (delays the effective date of SB24-205 requirements); Colorado General Assembly, SB26-189 Automated Decision-Making Technology. Download as image

AI hiring law status by state

• 4State law or regulation on AI in employment decisions
• 1City law only (New York City)
• 1Narrow rule (facial recognition in interviews)
• 1General AI law (intentional discrimination only)
• 44No AI-specific hiring law identified

Which states have AI hiring laws in 2026?

Four states have rules written for automated decisions about workers: California, Colorado, New Jersey and Illinois. New York has no statewide law, but New York City's Local Law 144 has been enforced since July 5, 2023. Maryland covers one narrow use, facial recognition during interviews, and Texas has a general AI law that bans only intentional discrimination.

Each works differently. NYC requires a bias audit within one year before you use a tool, publication of the results, and notice to candidates 10 business days before use. Under California's rules, an automated decision can count as discrimination under FEHA, and employers must keep the related records for at least 4 years. New Jersey's rules, in force since December 15, 2025, say these tools can cause unlawful disparate impact and flag tools not tested before use as a risk. Colorado's SB 26-189, signed May 14, 2026, requires notice when someone interacts with the system and an explanation within 30 days after an adverse outcome.

Illinois is the least certain entry on this page. Its Human Rights Act now covers AI in employment decisions and a separate act covers AI analysis of video interviews, but we could not confirm the effective dates or detailed duties from the statute text.

Does an AI hiring law apply if my company is in another state?

It can. The rules follow the job and the candidate more than your headquarters. NYC's law applies when the job is tied to a New York City office, even a fully remote one, and its notice goes to candidates who live in the city. California's rules protect applicants and employees in California. A small company in Ohio with one office in Queens and a remote hire in Los Angeles has picked up two sets of rules.

The simplest fix is to decide up front where you will hire, and to turn on notice and audit features for the strictest place on that list. Running one process for everyone is cheaper than tracking which rule applies to which résumé. If a vendor charges extra for audit reports or notice templates, count that in the price before you sign.

What happens if an employer breaks these rules?

Penalties vary a lot. In New York City, a first violation costs up to $500 and each later one up to $1,500. That sounds small until you remember that each candidate who did not get notice can count separately.

Texas sits at the other end. Under its new AI law, an uncurable violation can bring a civil penalty of $80,000 to $200,000, though the state has to show you meant to discriminate. Colorado gives employers 60 days to cure a problem before the attorney general can sue, at least until January 1, 2030. California leaves enforcement to the normal FEHA remedies, which means a discrimination claim with back pay and damages on the table.

• NYC: up to $500 first violation, up to $1,500 each later one
• Texas: up to $200,000 per uncurable violation
• Colorado: attorney general enforcement, with a 60-day cure notice before 2030
• California and New Jersey: the same remedies as any discrimination claim

Who is liable when the vendor's tool is biased?

Mostly you. Nearly all the rules on this page put the main duties on the employer that uses the tool. A vendor may run the model, but your name is on the rejection email.

That is why the vendor contract matters more than the sales deck. Many software agreements cap the vendor's liability at a few months of fees, disclaim any promise that the output is accurate or fair, and say nothing about audits. If a rejected applicant files a claim, a clause like that leaves you paying for a defect you could not see.

Ask for the vendor's most recent bias audit before you sign. If you hire for NYC-based jobs, you need one done within the past year anyway. Ask too what data trained the model and whether your applicants' data will be used to train it further.

What should a small employer do before using AI screening?

Start with a plain inventory. List every tool that scores, ranks, filters or interviews candidates, including features buried inside your applicant tracking system. You may be using AI in hiring without ever having bought anything labeled AI.

Then work through a short routine for each one:

• Find out where each job is based and where applicants live, and which rules from the table apply
• Get the vendor's latest bias audit and the date it was done
• Send notice before the tool is used, using NYC's 10 business days as the safe default
• Keep a person in the loop for rejections, and offer human review when someone asks
• Keep records of inputs and decisions for at least 4 years to meet California's rule
• Offer a clear way to request an accommodation, which New Jersey now expects for schedule screens

What this means for job seekers

If you live in New York City and apply for a job tied to a city office, you are entitled to notice before an automated tool assesses you, and the employer has to publish its audit results. In Colorado, if the decision goes against you, you can ask how the technology was used and ask for a human to look again.

Elsewhere your rights are thinner, but general discrimination law still applies. If you are rejected within minutes of applying, or an interview platform asks to scan your face, it is fair to ask the employer what the tool does. In Maryland, they need your signed consent for that face scan first.

Sources and method

Every number on this page comes from the sources below, each opened and read on or before October 18, 2026. A second, independent review checked every figure against its source and corrected anything that did not match. 51 of 51 state entries were confirmed; any that could not be are marked "Not yet confirmed" and left out of every count and chart.

1. New York City Council, Local Law 144 of 2021 (automated employment decision tools), legislation detail and text (Enacted December 11, 2021)
2. NYC Department of Consumer and Worker Protection, Automated Employment Decision Tools (AEDT) (Accessed October 2026; enforcement from July 5, 2023)
3. NYC Department of Consumer and Worker Protection, Automated Employment Decision Tools: Frequently Asked Questions (June 29, 2023)
4. California Civil Rights Department, Civil Rights Council Secures Approval for Regulations to Protect Against Employment Discrimination Related to Artificial Intelligence (June 30, 2025)
5. California Privacy Protection Agency, CCPA Updates, Cybersecurity Audits, Risk Assessments, Automated Decisionmaking Technology (ADMT), and Insurance Companies regulations (Approved by OAL September 22, 2025; effective January 1, 2026)
6. California Privacy Protection Agency, Text of Regulations: CCPA Updates, Cybersecurity Audits, Risk Assessments, ADMT and Insurance Companies (11 CCR § 7200) (Approved regulation text; effective January 1, 2026)
7. New Jersey Office of the Attorney General, Division on Civil Rights, N.J.A.C. 13:16, Disparate Impact Discrimination (R.2025 d.150) (Effective December 15, 2025)
8. Illinois Department of Human Rights, Legislative Updates (Accessed October 2026)
9. Maryland General Assembly, Maryland Code, Labor and Employment § 3-717 (facial recognition services in interviews) (Current code, accessed October 2026)
10. Texas Legislature, HB 149 enrolled text, Texas Responsible Artificial Intelligence Governance Act (89th Legislature, Regular Session, 2025)
11. Texas Legislature Online, HB 149 bill history (Signed June 22, 2025; effective January 1, 2026)
12. Colorado General Assembly, SB24-205 Consumer Protections for Artificial Intelligence (Approved by Governor May 17, 2024)
13. Colorado General Assembly, SB25B-004 (delays the effective date of SB24-205 requirements) (Approved by Governor August 28, 2025)
14. U.S. Census Bureau, Business Trends and Outlook Survey (BTOS) AI Supplement Table 2026 (national, state and employment size estimates) (2026 edition (BTOS 2025-2026 data))
15. Pew Research Center, AI in Hiring and Evaluating Workers: What Americans Think (April 20, 2023 (survey Dec 12-18, 2022))
16. Colorado General Assembly, SB26-189 Automated Decision-Making Technology (Governor signed May 14, 2026 (Session Law Ch. 131))