Your hiring AI is now a civil-rights question.

What the law does

HB 3773 amends the Illinois Human Rights Act, the state's core anti-discrimination statute, to address artificial intelligence directly. It took effect January 1, 2026, and it does two main things. First, it makes it a civil-rights violation for an employer to use AI that has the effect of subjecting employees to discrimination on the basis of a protected class — and notably, the violation can turn on discriminatory effect, not only intent, so an employer can run afoul of it without meaning to. Second, it requires employers to notify employees and applicants when AI is used in covered employment decisions. The law also addresses using zip codes as a proxy for protected classes. The amendment does not create a new standalone private right of action; instead it plugs into the Illinois Human Rights Act's existing enforcement machinery — a complainant files a charge with the Illinois Department of Human Rights, after which the matter proceeds before the Human Rights Commission or, at the complainant's election, in Illinois circuit court, with remedies available under the IHRA depending on forum and facts (which may include make-whole relief, civil penalties, and fees where authorized; verify against the current IHRA provisions). That existing framework, plus the IDHR's implementing rules, are what give the law its teeth. As this was written, the IDHR had published proposed implementing rules in the Illinois Register (May 15, 2026), opening a 45-day public-comment period with comments due June 29, 2026; the rules were proposed, not final, so the current status should be confirmed before reliance. Notably, unlike the New York City and Colorado regimes, HB 3773 does not itself mandate a formal bias audit or impact assessment. The amendment reaches employers with one or more employees in Illinois for twenty or more calendar weeks in a year, and the covered decisions sweep broadly across the employment lifecycle — recruitment, hiring, promotion, renewal, training selection, discipline, discharge, and the terms and conditions of employment. These are the load-bearing features; the exact statutory text and the implementing rules should be confirmed against current law.

Who is covered, and how broadly AI is defined

The reach is wide. Because HB 3773 works through the Illinois Human Rights Act, it generally extends to employers across the state, and the definition of “artificial intelligence” it operates under is broad — broad enough to capture tools many employers do not think of as “AI” at all: resume screeners, automated assessments, chatbots in the application flow, scheduling and performance-management software, and similar systems that influence employment decisions. The practical first step for most employers is simply discovering where AI is actually touching their employment decisions, because it is frequently in more places than HR realizes — often embedded inside vendor tools the employer never evaluated as AI. The precise scope and any exemptions should be verified against the current statute and rules.

The two obligations: don't discriminate, and disclose

Compliance has a substantive side and a procedural side. The substantive obligation is that the AI must not produce discriminatory effects against protected classes — which means an employer needs to actually understand what its AI tools do and how they affect outcomes, not just trust a vendor's assurance. The procedural obligation is notice: telling employees and applicants, in a way they can understand, when AI is being used in covered decisions. Meeting the notice requirement well requires understanding how the tool works in the first place, which loops back to the substantive side. Drafting those disclosures and the supporting internal policy is concrete, deliverable work — and getting ahead of it was the right move even before the final rules landed.

Why operational experience matters here

Evaluating whether an employment AI tool creates discriminatory effects is not a paperwork exercise — it requires understanding how these systems actually behave, where bias enters, and what a defensible oversight step looks like. This is exactly where building and governing a production AI system is an advantage rather than a talking point. Adam advises Illinois employers on HB 3773 from that operational vantage: he holds the AIGP credential in AI governance and governs a multi-agent system daily, so the analysis of an employer's hiring or performance tool comes from someone who understands the machinery, not only the statute. HB 3773 also sits alongside Illinois's earlier Artificial Intelligence Video Interview Act and BIPA, so an employer's AI tools often implicate more than one law at once.

What employers should do

A practical compliance path generally runs: inventory every place AI touches employment decisions (including inside vendor tools); assess each for discriminatory-effect risk; implement the required notices and an internal AI-use policy; document the human oversight and the basis for relying on each tool; and monitor for the final IDHR rules and any changes, since the regulatory picture was still settling. The firm handles this hourly, on a governance retainer, or as a defined flat-fee HB 3773 assessment. The goal is an employer who can show — if challenged — what its AI does, that it was disclosed, and that a human was accountable for the decisions.

If you hire across state lines, HB 3773 is not the only rule

An employer whose workforce or applicant pool crosses state lines faces parallel AI-in-employment rules, and HB 3773 is one piece of a larger patchwork. New York City's Local Law 144, in effect since 2023, requires an independent bias audit, a public summary of the results, and advance notice to candidates before an automated employment decision tool is used. California's privacy regulations reach automated decision-making technology used in employment, with among the most stringent such obligations in the country phasing in over time. Other states continue to move. These specifics change and must be verified against current law before you rely on them. As a practical risk posture, a multi-state employer is usually better served building its AI-hiring oversight to the most demanding standard in its actual footprint than maintaining a different posture state by state — and where compliance turns on advising under another state's law, the firm coordinates with employment counsel admitted there rather than reaching beyond where it is authorized to practice.

What usually goes wrong

The most common failure is the employer who does not know it is using covered AI at all — the discrimination and notice obligations attach to tools embedded in vendor hiring and HR software that the employer never identified as “AI,” so it is out of compliance without realizing a law applies. A close second is relying on a vendor's blanket assurance that a tool is “bias-free” or “compliant,” when the liability under the Human Rights Act sits with the employer, not the vendor, and turns on effect rather than the vendor's promise. The third is treating notice as a one-line add-on without understanding the tool well enough to describe it accurately — a disclosure that misstates what the AI does is its own exposure.

Related

• AI Governance Frameworks & Policy →
• BIPA Compliance →
• AI Employment Policies & Workforce Deployment →