AI hiring laws in 2026: what employers actually have to do

AI hiring laws are the rules that govern how employers can use automated tools to screen, assess, and select candidates. In 2026 they cluster around a few requirements: audit your tools for bias, tell candidates when AI is used, keep a human in the loop, and be able to explain the decision.

This guide is the practical version for the people who run hiring, not the legal-brief version, covering what NYC, Illinois, Colorado, and the EU actually require you to change.

Most content on this topic is legal analysis written for lawyers. That is useful, but it rarely tells a hiring team what to do differently on Monday. The honest throughline here: compliance is not a separate legal chore, it is mostly a function of whether your hiring tools are explainable and audited, which is what a defensible process should be anyway.

This is practical guidance, not legal advice. Confirm your obligations with counsel before you rely on any of it.

Key Takeaways

The common thread across AI hiring laws is the same four things: bias audits, candidate notice, human oversight, and explainability. Nail those and most jurisdictions are covered.
NYC Local Law 144 is the one with teeth today: if a tool substantially assists a hiring decision, it is an automated employment decision tool and needs a published bias audit and candidate notice.
AI hiring laws by state are multiplying (Illinois, Colorado, and more), so building compliance into your process beats chasing each new statute.
The EU AI Act classifies hiring AI as high-risk, which means real documentation and oversight obligations if you hire in Europe.
A black-box AI score you cannot explain is the core legal risk. Explainable, audited scoring is both the compliance answer and the hiring-quality answer.

What AI hiring laws are and why they exist

AI hiring laws (sometimes called ai employment laws) regulate automated tools used to recruit, screen, assess, or select candidates: resume screeners, video-interview analyzers, and AI scoring. They exist because these tools can reproduce bias at scale. A model trained on a company's past hiring can learn that company's past bias, and then apply it to thousands of candidates faster than any human panel could.

The laws respond to that risk in a consistent way. They require employers to check tools for discriminatory impact, disclose when AI is being used, and keep a human accountable for the decision. The details differ by jurisdiction, but the shape is the same, which is good news: you can build one compliant process rather than one per law.

NYC Local Law 144: the automated employment decision tool law

New York City's Local Law 144 is the most concrete automated employment decision tool law in force. If a tool substantially assists or replaces a hiring or promotion decision, it is an automated employment decision tool (an AEDT), and using it triggers two obligations.

First, a bias audit: an independent audit of the tool's impact across sex, race, and ethnicity, published on your site and refreshed annually. Second, candidate notice: you must tell candidates that an AEDT is being used, what characteristics it assesses, and give them a chance to request an alternative. Expert Hire's NYC Local Law 144 compliance page breaks down the specifics.

The practical takeaway: if you screen NYC candidates with any AI tool, you need audit documentation from your vendor and a notice in your process. If a vendor cannot produce a bias audit, that is your answer.

Illinois was early here. The Artificial Intelligence Video Interview Act (820 ILCS 42) governs employers who use AI to analyze video interviews. It requires you to notify applicants that AI will be used, explain how it works and what it evaluates, obtain consent before the interview, and delete the video within a set period on request.

The through-line with NYC is disclosure and consent. If you use AI to analyze recorded interviews for Illinois roles, your process needs an explicit consent step and a data-retention answer. Our Illinois AI Video Interview Act page covers the details.

Colorado and the wave of AI hiring laws by state

Colorado's AI Act (SB 24-205) is the broader model that other states are following. It treats employment as a high-risk use of AI and requires deployers to use reasonable care to avoid algorithmic discrimination, to disclose AI use to candidates, and to maintain documentation of risk management.

Colorado is not alone; ai hiring laws by state are multiplying, with more legislatures introducing AEDT-style bills each session. Chasing each new statute individually is a losing game. The winning move is to build the common requirements (audit, notice, oversight, documentation) into your process once, so a new state law is a checkbox rather than a fire drill.

The EU AI Act: hiring AI as high-risk

If you hire in Europe, the EU AI Act applies. It classifies AI systems used in recruitment and candidate selection as high-risk, which is the strictest tier short of prohibited. High-risk classification brings real obligations: risk management, data governance, human oversight, transparency to affected people, and technical documentation.

In plain terms, the EU expects you to know how your hiring AI works, to keep a human accountable, and to be able to show your documentation. It is the most demanding of the major regimes, and it rewards the same discipline the US laws do: explainability and oversight.

What you actually have to change in your hiring process

Strip away the jurisdictions and the practical ai hiring compliance checklist is short:

Get a bias audit for every AI tool that assists decisions.

Ask vendors for their documentation before you buy. No audit, no deal.
Add candidate notice and, where required, consent.

Tell candidates when AI is used and what it assesses; build a consent step for AI-analyzed video where the law requires it.
Keep a human accountable.

AI can triage and score, but a person should own the advance-or-reject decision, especially at the final stages.
Insist on explainable scoring.

You must be able to say why a candidate got the score they did. A score with no reasoning fails every one of these laws.
Keep documentation.

Retain audit results, vendor documentation, and your own risk notes.

Do these five things and you are compliant across most jurisdictions at once, because they are the shared core of nearly every one of these ai interview laws.

How to choose compliant hiring tools

When you evaluate a hiring tool, three questions settle most of the compliance picture. Can the vendor produce a current bias audit? Can the tool explain why it scored a candidate the way it did? Does it keep a human in the loop rather than auto-rejecting?

This is exactly why Expert Hire's scoring methodology is published openly: a score you cannot inspect is a score you cannot defend to a regulator or a candidate. The AI interview platform is built to run a structured, explainable round with a human owning the decision, and structured interview software makes the process consistent and auditable by design. Compliance and hiring quality point in the same direction here.

Frequently asked questions

What states have AI employment laws? Illinois (the AI Video Interview Act) and New York City (Local Law 144) have specific, enforceable rules today, and Colorado's AI Act brings a broader high-risk framework. More states introduce AEDT-style bills each session, so treat this as a growing list and build compliance into your process rather than tracking each law.

What is an automated employment decision tool? Under NYC Local Law 144, an automated employment decision tool (AEDT) is software that uses machine learning or AI to substantially assist or replace a hiring or promotion decision. If a tool meaningfully drives who advances, it is likely an AEDT and triggers the bias-audit and notice requirements.

Do I need a bias audit for my hiring AI? In New York City, yes, if the tool is an AEDT: an independent bias audit, published and refreshed annually. Even where not strictly required, a bias audit is the single best evidence that your tool is not producing adverse impact, so it is worth doing regardless.

Are AI hiring tools legal? Yes, when used within the rules: audited for bias, disclosed to candidates, kept under human oversight, and explainable. The laws regulate how you use AI in hiring; they do not ban it. The risk is not the tool, it is using a black-box tool with no audit or oversight.

Does the EU AI Act apply to US companies? It can. If you use AI to recruit or select candidates who are in the EU, the high-risk obligations can apply regardless of where your company is based. If you hire in Europe at all, assume it is in scope and confirm with counsel.

The bottom line

AI hiring laws look like a patchwork, but they rhyme: audit for bias, tell candidates, keep a human accountable, and be able to explain the decision. Build those four things into your hiring process and you are compliant across most jurisdictions without chasing each new statute. The pages that scare you into buying legal services are not wrong, but the fix is mostly operational, not legal.

The clearest signal that a tool will keep you compliant is whether it can explain itself. Look at how Expert Hire scores a candidate, the rubric, the reasoning, the human in the loop, and decide whether that is the kind of defensible, auditable process your hiring should run on. For the policy side of the same question, our take on whether you should allow AI in job interviews is the companion read.

By TK, Growth at Expert Hire. Last updated July 6, 2026. Reviewed by Anand Suresh, CPO at Expert Hire.