Illinois’ New AI Hiring Disclosure Law Is Live — Here’s the India Exposure Nobody’s Mapped Yet

Illinois House Bill 3773 took effect on January 1, 2026, and it does something most state AI-hiring laws haven’t done yet: it makes a discriminatory outcome from an AI screening tool a civil rights violation on its own, not just evidence in a broader discrimination case, and it requires employers to notify candidates whenever AI is used in an employment decision that affects them (Ogletree Deakins, “Illinois Unveils Draft Notice Rules on AI Use in Employment”). Employment lawyers have started calling Illinois a “plaintiff’s blueprint state,” because the law creates a direct civil right of action tied to both discriminatory AI use and to employers who skip the disclosure step (DarrowEverett, “AI Hiring & Workforce Management: 2026 Legal Analysis”).

If you’re running screening operations out of India, your first instinct is probably that this is a US HR problem. It isn’t, and the gap between those two facts is exactly where the exposure sits.

Why an Illinois state law reaches an India-based screening team

A large share of India’s screening volume isn’t for India-based roles. Global Capability Centers, IT services majors, and BPO/staffing firms screen candidates in India for positions that report into US business units — including remote-eligible roles that legally sit wherever the worker is employed. Illinois HB-3773’s obligations attach to the employment decision and the candidate, not to where the AI vendor’s servers or screening team happen to be. A recruiter in Pune running a first-round AI voice screen for a Chicago-headquartered client’s Illinois-based support role is inside the law’s scope the same way a recruiter physically sitting in Chicago would be.

California moved on a parallel track. Amended Fair Employment and Housing Act regulations took effect October 1, 2025, explicitly covering automated decision systems in employment, and starting October 1, 2026, using an automated tool stops being a valid defense against a discrimination claim — layoff notices will also have to disclose whether AI drove the decision (K&L Gates, “California Lays the Groundwork for More Sweeping AI Workforce Regulation”). Candidate-facing notice obligations follow in October 2027. Two of the largest US states with the heaviest concentration of GCCs and captive tech centers now have live or imminent AI-hiring disclosure regimes, and neither cares whether the screening happened in San Jose or Bengaluru.

What this actually requires, in practice

Strip away the legal language and three obligations repeat across Illinois, California, and — for a different reason — New York’s Local Law 144, which is heading into a stricter 2026 enforcement phase after a state Comptroller audit found the city’s own oversight body had been under-enforcing it, with misrouted complaints and superficial reviews of publicly posted bias audits (Office of the New York State Comptroller, enforcement audit): notice to the candidate that AI was used, documentation of how the tool evaluates candidates, and evidence you tested the tool for discriminatory outcomes rather than just trusting the vendor’s marketing claims.

None of that is exotic. It’s closer to what any structured, auditable screening process should already produce. A voice interview that generates a full transcript and applies the same evaluation criteria to every candidate is, by construction, easier to document than an unstructured phone call where the outcome depended on which recruiter picked up and how their day was going. We’ve written before about the four questions your legal team will actually ask once AI hiring gets flagged for review — disclosure and documented testing are now the specific, state-dated version of that same conversation.

The contrarian part

Here’s what most vendor sales decks won’t say out loud: patchwork state law is actually harder on screening tools that can’t explain their own decisions than it is on the underlying practice of AI screening itself. A tool that outputs a black-box score with no transcript, no stated criteria, and no consistent rubric across candidates is the thing these laws are designed to catch. A tool that produces a complete record of what was asked, what was said, and why a candidate landed in a given tier isn’t fighting the regulation — it’s already halfway to compliant. The exposure isn’t “we use AI to screen.” It’s “we can’t show our work.”

What to do before your next US-facing campaign

If your India team screens candidates for any US client with Illinois or California-based roles — including fully remote ones — three things are worth doing this quarter, not after a complaint arrives. First, ask your AI screening vendor, ours included, whether candidates are told AI is part of the process before the interview starts, in plain language. Second, confirm you can produce a transcript and a stated evaluation rubric for any candidate who disputes an outcome. Third, check whether your vendor has actually tested for discriminatory outcomes across demographic groups, versus simply asserting the tool is fair.

The compliance clock on Illinois HB-3773 is already running. California’s is six months out. Mapping this exposure now, while it’s still a documentation exercise, is considerably cheaper than mapping it after the first complaint names your screening vendor. If you want to see what an auditable, transcript-backed screening rubric looks like on a live campaign, book a demo.