Table of Contents
Quick Answer
Using AI in hiring, promotion, and workforce management in 2026 requires compliance with anti-discrimination statutes (Title VII, ADA, ADEA, Equality Act 2010, Indian Constitution Art. 14-16) and AI-specific rules (NYC Local Law 144, Illinois AI Video Interview Act, Colorado AI Act, EU AI Act Annex III).
- EEOC, DOJ, OFCCP, and CFPB all police AI-assisted employment decisions in the US
- NYC Local Law 144 is the template for AI bias audits in hiring
- EU AI Act classifies recruitment AI as high-risk (Annex III point 4)
What Is AI Employment Law?
AI employment law is the intersection of traditional anti-discrimination law and AI-specific regulation governing the use of automated employment decision tools (AEDTs). The EEOC's May 2023 technical assistance document and its 2024 ADA AI guidance are the federal US baseline; state and city laws add bias audit and notice obligations.
Key Details / Requirements
US State and City AI Employment Laws
Law
Scope
Core Obligation
NYC Local Law 144 (effective 2023)
AEDTs for NYC jobs
Independent bias audit, candidate notice, annual publication
Illinois AI Video Interview Act (2020)
Video interview AI
Notice, consent, report to IDOL
Maryland HB 1202 (2020)
Facial recognition in hiring
Written consent
Colorado AI Act (2026)
High-risk AI including hiring
Impact assessment, notice, AG filing
California SB 7 (proposed)
Workplace AI
Notice, risk assessment
NYC Local Law 20 of 2025
Workplace surveillance AI
Advance notice
EU AI Act Employment Obligations (Annex III point 4)
Role
Obligation
Provider
Conformity assessment, CE marking, technical file
Deployer
Fundamental Rights Impact Assessment, human oversight, worker representative consultation
Both
Transparency to affected workers
Real-World Examples / Case Studies
iTutorGroup (2023) — Paid USD 365,000 in the first EEOC AI-hiring discrimination settlement after software screened out women 55+ and men 60+.
Workday (N.D. Cal., pending) — Proposed class action asserts disparate impact on older applicants.
HireVue — Removed facial analysis from its interview assessments (January 2021) after civil-rights complaints.
Amazon scrapped resume tool (2018) — Internal tool down-weighted resumes with "women's" keywords; project cancelled.
Uber drivers (UK) — Aslam v. Uber (Supreme Court, 2021) held drivers were workers, relevant to AI-managed gig-work claims.
What This Means for Employers
In 2026, employers using AI in HR must:
- Run pre-deployment bias audits (NYC 144, Colorado AI Act)
- Provide candidate notice 10 business days before use of AEDT
- Offer reasonable accommodation (ADA) — AI must not screen out disabled applicants
- Keep records for adverse-action defence (4/5ths analysis)
- Consult with worker representatives where required (EU AI Act Art. 26(7))
- Train hiring managers on AI limitations
Compliance Checklist
- Inventory every AEDT and classify under applicable laws
- Commission an annual independent bias audit (NYC Local Law 144)
- Publish candidate-facing AEDT notices
- Provide alternative processes for disabled applicants (ADA)
- Document reasonable-accommodation workflows
- File Colorado AI Act impact assessments for high-risk systems
- Align with EU AI Act Annex III and Art. 26 for EU workforces
- Update employee handbooks to cover AI use
FAQs
Q: Does Title VII apply to AI?
Yes — disparate-treatment and disparate-impact doctrines apply to AI-driven hiring and promotion decisions.
Q: What is an AEDT?
Automated Employment Decision Tool — defined in NYC Local Law 144 as a computational process that substantially assists or replaces discretionary decision-making.
Q: Does the ADA cover AI?
Yes — EEOC's 2024 guidance explains that AI screening must not disadvantage disabled applicants and must provide reasonable accommodation.
Q: Are bias audits mandatory?
For NYC AEDTs and Colorado high-risk hiring AI, yes.
Q: What is the 4/5ths rule?
EEOC's rule-of-thumb: selection rate for any group must be at least 80% of the highest rate.
Q: How do EU works councils factor in?
EU AI Act Art. 26(7) requires informing worker representatives before deploying high-risk AI at work.
Q: Can I be sued for using a third-party AI tool?
Yes — deployers are liable under Title VII; indemnification from the vendor can shift but not eliminate risk.
Conclusion
AI in HR is one of the most regulated frontiers. Employers who treat bias audits as routine avoid both fines and headlines.
Stay compliant with Misar AI's HR AI governance toolkit — NYC Local Law 144 ready.