Here is a question we ask HR and payroll leaders a lot: how many AI tools touch your hiring process today?

The honest answer is almost always "I'm not sure."

Resume screening in the ATS. Candidate ranking in the HCM. Interview scheduling. Assessment scoring. A lot of it got switched on quietly in a product update.

And now it is regulated. Not someday. Now.

#1. Illinois has been live since January 1, 2026

Illinois amended its Human Rights Act with HB 3773 (Public Act 103-0804). If you use AI in recruiting, hiring, promotion, discipline, discharge, or other terms of employment:

• You have to tell people. Applicants and employees get notice.

• Intent does not matter. If the AI discriminates against a protected class, that is a civil rights violation, even if nobody meant it to.

• No zip codes as a proxy. Explicitly called out.

The state has proposed detailed rules on what the notice says and where it goes, including job postings and handbooks. Those details may still move. The law itself is not waiting.

#2. Colorado starts January 1, 2027, and it's narrower than you heard

Ok, ok, you say. Didn't Colorado pass some huge AI law?

It did. Then in May 2026 it replaced it with SB 26-189, a much tighter law on automated decision-making technology. For employers, it covers tech that materially influences hiring, pay, promotion, and similar decisions. Clerical uses are out.

• Notice before you use it, for example on the job posting or careers page.

• An explanation within 30 days when someone gets an adverse outcome.

• A right to fix bad data and to ask for human review when that is commercially reasonable.

• Three years of records.

Only the Attorney General enforces it. No private lawsuits. But the AG is writing rules, and they will add detail.

#3. California got there first

California's civil rights regulations on automated decision systems took effect October 1, 2025. The core idea is simple: if your hiring software filters people out unfairly, that is discrimination. The algorithm does not get you off the hook.

#4. What to actually do this quarter

You do not need a 40-page AI policy. You need five things:

1. Inventory every tool that screens, scores, or ranks people. Include features built into your ATS and HCM.

2. Ask your vendors what the tool does, what data it uses, and how it was tested for bias. If they cannot answer, that is your answer.

3. Write your notices for job postings, the careers page, and the handbook.

4. Put a human on adverse decisions, and log who reviewed what.

5. Map where your people are. These rules follow where applicants and employees live and work, not just your HQ.

The bottom line

The pattern across every state is the same. Tell people when AI is involved. Keep records. Keep a human in the loop.

That is how we built Tada. We use AI for the back office: reading timesheets, processing payroll changes, checking compliance. Your team reviews and approves. Every change keeps an audit trail. It is the same bar these laws now set for hiring tools, and it is the right one.

Explore Tada products or book a demo.

General information, not legal advice. Talk to employment counsel about your specific situation.