Illinois changed its Human Rights Act so that using artificial intelligence in a way that discriminates is a civil rights violation, and so is failing to tell an employee that AI is being used. The change came from House Bill 3773 and took effect on 1 January 2026.
The statute left the details of the notice to the Illinois Department of Human Rights. The Department proposed rules in May 2026 and withdrew them in June. This article sets out what the statute says, what the withdrawn rules would have required, and what is still open.
What does the Illinois AI hiring law say?
It adds a new subdivision to the employment article of the Illinois Human Rights Act that covers discriminatory use of AI and failure to give notice of AI use.
The enrolled text of Public Act 103-0804 adds subdivision (L), titled 'Use of artificial intelligence', to section 2-102 of the Act. Paragraph (1) covers an employer that uses artificial intelligence 'that has the effect of subjecting employees to discrimination on the basis of protected classes' in a listed set of employment decisions, or that uses zip codes as a proxy for protected classes. Paragraph (2) covers an employer that fails to provide notice to an employee that it is using artificial intelligence for those purposes.
The same subdivision tells the Department of Human Rights to adopt any rules necessary for implementation and enforcement, including rules on 'the circumstances and conditions that require notice, the time period for providing notice, and the means for providing notice'. Section 99 of the Act sets the effective date at 1 January 2026.
The wording of paragraph (1) turns on effect. The text does not ask whether the employer meant to discriminate. Read the statute itself for the exact language, because short summaries tend to drop that point.
| Topic | Short answer | Source |
|---|---|---|
| Law | Public Act 103-0804, from House Bill 3773, amending the Illinois Human Rights Act | Illinois General Assembly |
| Effective date | 1 January 2026 | Public Act 103-0804, section 99 |
| First violation | Using AI that has the effect of discrimination on a protected basis, or using zip codes as a proxy | 775 ILCS 5/2-102(L)(1) |
| Second violation | Failing to give notice that AI is used for the listed decisions | 775 ILCS 5/2-102(L)(2) |
| Notice rules | Proposed 15 May 2026, withdrawn 2 June 2026 | Burke Warren alert, 16 June 2026 |
| Bias audit | Not required by the statute | Hinshaw, as cited in the research notes |
| Time to file a charge | Within 2 years of the alleged violation | 775 ILCS 5/7A-102 |
Who is covered and which decisions count?
Employers as the Human Rights Act defines them, for decisions from recruitment through discharge.
The Act's definition of employer includes any person employing one or more employees within Illinois during 20 or more calendar weeks in the calendar year of or before the alleged violation. It also includes the State, its political subdivisions and parties to public contracts regardless of headcount. The threshold is in section 2-101 of the Act.
Subdivision (L) lists the decisions it reaches: recruitment, hiring, promotion, renewal of employment, selection for training or apprenticeship, discharge, discipline, tenure, and the terms, privileges or conditions of employment. That list is wider than hiring. A scoring tool used in performance reviews or discipline falls within the same words.
Hinshaw's February 2026 summary says the draft rules also reached agents of an employer, such as recruiters and other third parties acting for it. That point comes from a draft that was later withdrawn, so treat it as a sign of the Department's thinking and check the current text.
What counts as artificial intelligence under the Act?
A machine-based system that infers from its inputs how to generate outputs such as predictions, content, recommendations or decisions.
Public Act 103-0804 added two definitions to section 2-101. 'Artificial intelligence' means 'a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments'. The definition says it includes generative artificial intelligence.
'Generative artificial intelligence' is defined as an automated computing system that, when prompted, can produce outputs that simulate human-produced content. The Act gives text, images and multimedia as examples.
The definition is broad. It does not name products, and it does not limit itself to tools sold as hiring software. Covington's June 2026 summary of the proposed rules said notice would have been required where AI was used to 'influence or facilitate' a covered decision, and that uses 'merely incidental' to a decision generally would not have needed notice. Hinshaw gave examples of routine software the draft treated as outside the notice duty, such as word processors, spreadsheets and spam filters. Those lines were in the withdrawn text and are not law.
What notice is required, and to whom?
The statute requires notice to an employee that AI is being used for the listed purposes, and leaves content, timing and method to rules that do not yet exist.
Paragraph (2) is one sentence. It refers to notice to 'an employee'. It does not say what the notice must contain, when it must be given or how.
The law-firm summaries read the duty as reaching applicants too. Burke Warren describes the statutory obligation as notifying applicants and employees when AI is used in employment decisions. The proposed rules would have handled applicants through the job posting, according to Covington. Whether paragraph (2) by itself reaches applicants is a question for counsel, since the sentence uses the word employee.
The table shows what the withdrawn proposal would have required, as reported by the three firms. It is included because it is the only public indication of what the Department considered adequate.
| Item | Withdrawn proposal, as reported | Reported by |
|---|---|---|
| Content | The AI product name, developer and vendor; the decision it influences; the job positions it is used for; a contact for questions; the right to request a reasonable accommodation | Burke Warren; Covington |
| Purpose and data | The system's purpose and the categories of data it processes | Burke Warren; Hinshaw |
| Timing for current employees | Annually, and within 30 days of adopting a new or substantially updated AI system | Hinshaw; Covington |
| Timing for applicants | In the job notice or posting | Covington |
| Where | Handbooks or policy documents, places where notices are customarily posted, and job postings | Hinshaw; Covington |
| Accessibility | Plain language, languages commonly spoken by the workforce, accessible to employees with disabilities | Hinshaw; Covington |
| Record retention | Three years in the May 2026 proposal according to Covington; four years in the earlier draft according to Hinshaw | Covington; Hinshaw |
What happened to the notice rules?
The Department published proposed rules on 15 May 2026 and withdrew them on 2 June 2026 without giving a new timeline.
Hinshaw reported in February 2026 that the Department had released draft rules titled 'Subpart J: Use of Artificial Intelligence in Employment'. Burke Warren reports that formal proposed rules followed on 15 May 2026, that the Department withdrew and postponed them on 2 June 2026, and that the public hearing set for 10 June 2026 was canceled. Burke Warren's alert gives no reason for the withdrawal.
Burke Warren describes the withdrawal as a pause in rulemaking and says the statutory obligations stay in force in the meantime. The research notes behind this article record one source saying the rules had not been refiled as of late September 2026. A search for this article on 12 November 2026 found no refiled or adopted rule. The Department's own page on the topic could not be opened, so that status rests on law-firm reporting.
The two summaries of the retention period do not match. Hinshaw's February article says four years. Covington's June article on the proposed rules says three years after the use. They describe different versions of the text. Neither is in force.
How is the law enforced?
Through a charge filed with the Department of Human Rights, with relief ordered by the Human Rights Commission.
Subdivision (L) sits in the list of civil rights violations in employment, so it uses the Act's existing process. Section 7A-102 says a written charge may be filed with the Department within 2 years after the date a civil rights violation allegedly was committed. The Department can also issue a charge itself.
Section 8A-104 lists the relief the Commission may order on finding a violation. It includes a cease and desist order, actual damages, hiring or reinstatement with or without back pay, attorney fees and costs, a compliance report, and action needed to make the complainant whole.
The research notes record that no enforcement action under HB 3773 was found, and they add that absence from search results does not prove there are none.
How does Illinois compare with other AI hiring rules?
Illinois has a notice duty and a discrimination rule but no audit duty, which places it between New York City and the notice-based regimes in Colorado and California.
The comparison below is drawn from the research notes and the sources they cite. Each of these regimes changed in 2025 or 2026, so the dates matter.
The practical difference is what an employer has to be able to show. New York City asks for an audit. Colorado and California ask for notices and records. Illinois asks for a notice whose required form has not been set.
| Jurisdiction | Status | Core duty | Source |
|---|---|---|---|
| Illinois, HB 3773 | In force since 1 January 2026. Notice rules withdrawn | No discriminatory effect from AI. Notice of AI use | Public Act 103-0804; Burke Warren |
| New York City, Local Law 144 | In force since July 2023 | Independent bias audit of automated hiring tools. A December 2025 state audit called enforcement ineffective | New York State Comptroller |
| Colorado, SB 26-189 | Effective 1 January 2027 | Pre-use notice, adverse-outcome notice within 30 days, human review, records for three years | Davis Wright Tremaine |
| California, Civil Rights Council regulations | In force since 1 October 2025 | Automated-decision records kept for four years | Mayer Brown |
Does the AI Video Interview Act still apply?
Yes. It is a separate statute from 2020 with its own notice, explanation and consent steps for recorded video interviews.
The Artificial Intelligence Video Interview Act took effect on 1 January 2020. It applies to an employer that asks applicants to record video interviews and uses an artificial intelligence analysis of those videos when considering applicants for positions based in Illinois.
Before asking for the video, the employer must notify the applicant that AI may be used to analyze the interview, explain how the AI works and what general types of characteristics it uses, and obtain the applicant's consent. The Act says an employer may not use AI to evaluate applicants who have not consented.
It also limits sharing of the videos and gives applicants a deletion right. On request, the employer must delete the interviews within 30 days and instruct anyone who received copies to delete them. An employer that uses AI video analysis has both statutes to read.
Records worth keeping while the rules are unsettled
A list of the tools in use, the notices given, who received them and when, kept in a form that can be produced later.
The statute makes failure to give notice a violation today, and a charge can be filed up to two years after the event. The employer will be the one asked to show that a notice was given. None of the items below is prescribed by the statute. They follow from the questions a charge would raise and from the content of the withdrawn proposal.
This is a description of what the sources say, not advice on what your business must do. An employment lawyer can say how the Act applies to a particular tool.
- An inventory of each system that produces a score, ranking, recommendation or decision about applicants or employees, with its vendor.
- For each system, the decisions it feeds: screening, interview selection, promotion, discipline, scheduling or pay.
- The exact text of each notice, with its version and the date it came into use.
- For each person, which notice they received, by what means and on what date.
- Whether zip code or another location field is an input to any model, and why.
- Any accommodation requests and questions received after a notice, and the response.
- Any testing the vendor or the employer has done for disparate results, and the date.
Common questions
When did the Illinois AI hiring law take effect?
On 1 January 2026. Section 99 of Public Act 103-0804 sets that date. The Act amended the Illinois Human Rights Act, so it applies through that law and its charge process.
Does Illinois require a bias audit of hiring software?
The statute does not mention an audit or an impact assessment. The Hinshaw summary cited in the research notes reads it the same way. New York City is the jurisdiction in this comparison that requires an independent bias audit. An employer may still choose to test a tool, because the Illinois violation turns on discriminatory effect.
What must an Illinois AI notice say?
The statute does not say. It leaves the circumstances, timing and means of notice to Department rules. Proposed rules listed items such as the product name and vendor, the decision affected, a contact point and the right to request an accommodation, but those rules were withdrawn on 2 June 2026 according to Burke Warren.
Does the law apply to small employers?
The Human Rights Act defines an employer to include any person employing one or more employees within Illinois during 20 or more calendar weeks in the relevant year. That is a low threshold. Check section 2-101 for the full definition and its exceptions.
Can an applicant sue under HB 3773?
The Act works through a charge filed with the Department of Human Rights within two years of the alleged violation, with relief ordered by the Human Rights Commission. How a claim moves from the Department to the Commission or to court is set out in the Act and is a question for an employment lawyer. This article does not cover that procedure.
Is using zip codes in a hiring model illegal in Illinois?
The statute makes it a civil rights violation for an employer to 'use zip codes as a proxy for protected classes' in the listed decisions. The text is about use as a proxy. Whether a particular use of location data is a proxy depends on facts this article cannot assess.
What is not known about this law?
Three things. The required content, timing and method of the notice are undefined because the rules were withdrawn. The research found no enforcement action, which may mean none exists or that none was reported. And the current rulemaking status could not be confirmed from the Department itself, only from law-firm reports.
Reviewed November 13, 2026
Consentra is software. Nothing on this site is legal advice, and using Consentra does not by itself make a business compliant with any law or carrier rule. Whether a rule applies to your business is a question for your counsel.
Sources
- Illinois General Assembly: Public Act 103-0804, House Bill 3773 (effective 1 January 2026)
- Illinois General Assembly: Illinois Human Rights Act, 775 ILCS 5/2-101, definitions
- Illinois General Assembly: Illinois Human Rights Act, 775 ILCS 5/7A-102, charge procedure
- Illinois General Assembly: Illinois Human Rights Act, 775 ILCS 5/8A-104, relief and penalties
- Illinois General Assembly: Artificial Intelligence Video Interview Act, 820 ILCS 42 (effective 1 January 2020)
- Hinshaw: Illinois adopts AI-in-employment regulations, what employers must know for 2026 (26 February 2026)
- Burke, Warren, MacKay & Serritella: Illinois Department of Human Rights withdraws proposed AI in employment rules (16 June 2026)
- Covington, Global Policy Watch: Illinois Department of Human Rights seeks public comment on draft AI employment regulations (10 June 2026)
- New York State Comptroller: Enforcement of Local Law 144, automated employment decision tools (December 2025)
- Davis Wright Tremaine: Colorado AI Act repeal and new transparency law (May 2026)
- Mayer Brown: California adopts new employment AI regulations effective 1 October 2025 (August 2025)
