AI in Recruiting: The Compliance Minefield

EEOC, NYC Local Law 144, and Colorado SB 205 all touch hiring AI. Here's what actually matters.

By Creative Genius · May 12, 2026 · 8 min read

If you use AI to screen resumes, score interviews, or rank candidates, you're operating in a compliance environment that changes every quarter. Get it wrong and you're not just exposed to fines — you're exposed to class-action lawsuits.

The three rules that actually have teeth

NYC Local Law 144 — requires a published bias audit of any "automated employment decision tool" used on NYC candidates, plus candidate notice. Effective since 2023, enforcement is real.
Colorado SB 205 — broad "high-risk AI" law including hiring. Requires impact assessments, risk management, and consumer notice. Phasing in through 2026.
EEOC enforcement guidance — Title VII applies to AI-driven hiring exactly as it applies to humans. Disparate impact is still illegal even when the model "didn't mean to."

The practical compliance stack

Document every model in use — version, training data, what input it sees, what output it produces, who it touches.
Pair every AI screen with a documented human-in-the-loop step. Auto-rejection without human review is a lawsuit waiting to happen.
Log everything candidate-facing — inputs, outputs, decision rationale. Six-year retention.
Budget for an annual third-party bias audit. $20K–$80K depending on volume. Treat it as a line item, not a project.
Candidate notice in plain language. "We use AI to assist with X. You may request review by a human."

What's safe to automate

Sourcing (suggesting candidates), scheduling, generating job descriptions, summarizing interview notes, drafting offer letters. None of these are "automated employment decisions" — they assist humans without making the call.

What's risky to automate

Resume scoring with auto-reject thresholds, video interview analysis used to rank candidates, personality assessments fed into hiring decisions. These are the regulated category. Document, audit, get sign-off from employment counsel.

Bottom line

AI in hiring is a real productivity unlock, but the regulatory floor is non-trivial. Build the documentation and audit cadence into the project from day one, or build a problem that takes years to unwind.