Job hunters in Illinois get new protection against AI hiring discrimination

AI hiring tools now face new limits

Companies across the country use artificial intelligence to screen resumes, rank candidates, and make hiring decisions. But these tools can pick up on patterns tied to where someone lives.

Because American neighborhoods are still heavily split by race and income, thanks to decades of housing discrimination, a zip code can say a lot about a person’s background.

Illinois passed a law that took effect Jan. 1, 2026, banning employers from letting AI use zip codes as a stand-in for protected traits like race, age, or religion.

The law covers every stage of hiring

House Bill 3773 amends the Illinois Human Rights Act.

It makes it a civil rights violation for employers to use AI that ends up discriminating against workers or job applicants based on any protected class.

The law separately bans using zip codes as a proxy for those traits. It covers everything from recruitment and hiring to promotion, training, discipline, and firing.

Any employer with at least one employee in Illinois falls under the law.

Employers must disclose AI use now

The law requires employers to tell workers and job applicants whenever AI plays a role in employment decisions. The Illinois Department of Human Rights handles the rules for how and when that notice goes out.

The department released draft notice rules in late 2025, though they had not yet gone through formal public comment.

Those draft rules define AI use broadly, covering any time an AI system’s output influences or helps make a decision about someone’s job.

Neutral data can still cause discrimination

Proxy discrimination happens when a neutral piece of information, like a zip code, stands in for a protected trait like race. An AI tool does not need to ask about someone’s race to discriminate.

It can learn that certain zip codes, schools, or employment gaps line up with certain groups. This kind of bias can happen without anyone meaning it to.

Under the Illinois law, if an AI system’s decisions end up hurting people of a certain race, age, or other protected group at higher rates, that counts as a violation.

Redlining still shapes where people live

The connection between zip codes and race goes back almost a century.

In the 1930s, the federal Home Owners’ Loan Corporation drew color-coded maps that graded neighborhoods by risk.

Areas with Black, Hispanic, and immigrant residents got marked as hazardous, which blocked people there from getting home loans.

According to the National Community Reinvestment Coalition, more than 8 million people still live in formerly redlined neighborhoods, and over three-quarters of those residents belong to minority groups.

Research from the Brookings Institution found that in cities like Atlanta and Minneapolis, average credit scores in mostly white neighborhoods top 725, while non-white zip codes average around 560 to 570.

Amazon learned this lesson the hard way

AI hiring tools learn from past hiring data, and if that history includes bias, the AI repeats it. Amazon built a resume-screening tool trained on 10 years of submitted resumes.

Most came from men, so the tool learned to penalize resumes with terms linked to women. Amazon scrapped it after finding the bias, but the case became a well-known warning.

Experts at the Urban Institute have noted that AI can learn to use graduation year as a proxy for age or zip code as a proxy for race, even when those traits are not in the data.

Lawsuits over AI hiring are growing fast

The legal challenges are stacking up. In Michigan, a job applicant claims Sirius XM’s AI screening system scored candidates based on home zip code and schools attended, which he says served as proxies for race.

In California, a judge let a case against hiring software company Workday move forward over claims its AI tools discriminated by age, race, and disability.

The Equal Employment Opportunity Commission settled its first AI discrimination case in 2023 after a tutoring company’s software automatically rejected women over 55 and men over 60, resulting in a $365,000 settlement.

Other states are writing their own rules

Illinois joins New York City and Colorado among the first places to directly regulate AI in hiring. New York City requires yearly independent bias audits of automated hiring tools.

Colorado’s AI law takes effect June 30, 2026, and requires impact assessments and transparency reports.

California’s Civil Rights Council finalized rules in 2025 that apply existing anti-discrimination law to automated decision-making.

Unlike New York City and Colorado, Illinois does not require formal bias audits, but legal experts say running them voluntarily could help employers defend against complaints.

Workers can file complaints at no cost

Workers or job applicants who think AI discriminated against them can file a complaint with the Illinois Department of Human Rights. The department then has 100 days to decide if there is enough evidence.

If it finds a violation, the case goes to the Illinois Human Rights Commission. Either side can ask to move the case to circuit court within 20 days.

Penalties under the Illinois Human Rights Act can include actual damages, attorney fees, and compliance reporting requirements.

Federal pushback adds uncertainty

The law passed the Illinois House 106-0 and the Senate 57-0 in 2024, and Gov. J.B. Pritzker signed it that August.

But in December 2025, President Trump signed an executive order directing the Department of Justice to challenge state AI laws, arguing regulation should stay at the federal level.

Legal experts say an executive order alone typically cannot override state law and that Congress would need to act to preempt laws like this one.

Separately, the Trump administration told the EEOC to deprioritize enforcement of disparate impact claims, though state enforcement stays intact.

The math behind catching AI bias

Federal guidance uses what is called the four-fifths rule to spot possible discrimination.

Here is how it works: if a hiring tool selects 60% of white applicants but only 30% of Black applicants, that ratio comes to 50%, well below the 80% threshold.

When one group’s selection rate falls below 80% of another group’s rate, regulators may treat that as evidence of discrimination.

This math applies whether a human or an AI does the screening, which is why employers face legal risk if they do not monitor their AI systems.

What workers and applicants should know

If you work or apply for jobs in Illinois, employers now have to tell you when AI plays a role in job decisions.

The law does not allow private lawsuits, but you can file a complaint with the Illinois Department of Human Rights at no cost. Workers in other states should keep an eye out too.

Colorado, California, and New York City have similar rules in place or on the way.

Federal anti-discrimination laws like Title VII still apply to AI hiring decisions nationwide, no matter what happens with state laws.

This article was created with AI assistance and human editing.

Read more from this brand:

• Grave #217 at this Illinois hospital tells a harrowing story
• RIP to free tiny shampoo bottles: Illinois hotels now face fines up to $1,500
• Japan can’t stop praising this garden in Illinois – here’s where to find it