In Illinois, AI in hiring is now a civil-rights issue. And it’s live.

Illinois did not pass a sweeping “AI Act.” It did something more pointed: it amended its civil-rights law. HB 3773 — signed August 9, 2024, and in force since January 1, 2026 — folds AI directly into the Illinois Human Rights Act. If you use AI in hiring and it discriminates, that is now a civil-rights violation in Illinois. No new agency, no new lawsuit type — the existing machinery just absorbed AI.

What changed

HB 3773 makes it a civil-rights violation for an employer to use AI — including generative AI — in a way that has the effect of discriminating against a protected class in employment decisions. The word that matters is effect. This reaches unintentional, disparate-impact discrimination, not just deliberate bias. A screening tool that quietly filters out a protected group can violate the law even if no one intended it to. Separately, the law bans using ZIP codes as a proxy for a protected class — a direct response to how location data can stand in for race or national origin.

The disparate-impact framing is what makes this law sharper than the intent-based statutes appearing elsewhere. You do not get to point to good intentions. If the tool’s outputs skew against a protected class, the burden lands on you to explain why — and “the vendor built it, not us” is not a defense. Under the Human Rights Act, the employer who uses the tool is the one accountable for the outcome. That single fact reframes vendor selection from a procurement decision into a compliance decision.

Which decisions are covered

The duty is not limited to the initial hire. It spans the full employment lifecycle:

• Recruitment and hiring.
• Promotion and renewal of employment.
• Selection for training or apprenticeship.
• Discharge, discipline, and tenure.
• The terms, privileges, and conditions of employment.

If AI touches any of those decisions — resume screeners, video-interview analysis, ranking algorithms, scheduling or performance tools — it is in scope. Many employers underestimate this footprint because the AI is buried inside familiar HR software. A modern applicant-tracking system may rank or auto-reject candidates with a model the employer never thinks of as “AI.” A performance platform may surface flight-risk or promotion scores. A scheduling tool may allocate shifts in ways that correlate with protected characteristics. Each of these can fall within the covered decisions, which is exactly why a tool-by-tool inventory — not a gut sense of “we don’t really use AI” — is the only reliable way to scope your exposure.

You also have to tell people

Beyond the anti-discrimination duty, HB 3773 imposes a notice obligation: employers must notify employees and applicants when AI is used for the covered purposes. The Illinois Department of Human Rights (IDHR) is adopting implementing rules — “Subpart J: Use of Artificial Intelligence in Employment” — that spell out the circumstances, timing, and means of that notice. Draft rules circulated in late 2025; verify the final text before relying on specifics, because the operational details live in those rules, not the statute itself.

How it’s enforced

This runs through the IDHR as a civil-rights charge under the Human Rights Act — the same administrative process that handles other discrimination complaints. That matters practically: it means an aggrieved applicant or employee files a charge, and your organization responds within an established framework with established remedies. It is not a novel, untested cause of action; it is a well-worn channel that AI has now been pulled into. Note too that Illinois’s older Artificial Intelligence Video Interview Act still applies separately to AI-analyzed video interviews; HB 3773 sits alongside it, not in place of it, so a single video-interview tool can implicate both. The exposure here is an employment-law exposure, and your vendor tools are the main risk surface — most employers do not build their own hiring AI, they buy it.

What “notice” really demands

The notice duty sounds simple and is not. To tell an applicant that AI is being used, you first have to know precisely where in your process it is used, what it does, and who is responsible for it. A vague banner that says “we may use automated tools” is unlikely to satisfy a rule built to give people meaningful awareness. The IDHR’s Subpart J is expected to define the timing (before the AI is applied, not after a rejection), the means (how the notice is delivered), and the circumstances that trigger it. Until that text is final, the defensible move is to over-document: record which tools you use, at which stage, and draft notice language you can refine once the rules settle. Building the inventory now means the notice is a small edit later, not a scramble.

Why this maps onto the same four duties

For all its civil-rights framing, HB 3773 asks for the same things every serious AI law converges on: anti-discrimination and notice. And both are impossible to satisfy if you cannot answer a basic question — which AI tools do we run in hiring, and how do they decide? You cannot give applicants notice about a system you have not catalogued, and you cannot defend against a disparate-impact charge without knowing what the tool weighed. The duty assumes an inventory you may not have.

It is worth saying plainly: this is the same skeleton that Colorado, Texas, and the EU all rest on, just dressed in Illinois civil-rights clothing. An employer that builds a clean inventory of its hiring AI, tests those tools for skewed outcomes, documents the testing, and notifies applicants has satisfied the spirit of HB 3773 — and, not by accident, most of what the other regimes ask for too. Build it once, and Illinois becomes one configuration of a portable program rather than a fire drill of its own.

Illinois shows where employment-AI regulation is heading: not a separate AI statute, but AI woven into existing civil-rights law — with unintentional harm squarely in scope. It is already live. The defensible position starts with knowing exactly which tools touch your hiring decisions.

This briefing is general information from Sentinel Assurance Group, not legal advice. Regulatory dates and requirements change — we maintain these briefings, but verify against primary sources and counsel before acting. Last reviewed June 11, 2026.

See how a Gap Assessment maps your exposure →