AI in hiring and HR is among the most enforced categories worldwide. Tools that screen résumés, score interviews, or rank candidates face mandatory audits, civil lawsuits, and regulatory action when they produce discriminatory outcomes — even unintentionally.
New York City Local Law 144
Since 2023, NYC requires annual independent bias audits for automated employment decision tools (AEDTs) used in hiring or promotion. Employers must publish audit summaries and notify candidates that AI is involved. Failure exposes employers to civil penalties — the law targets deployers, not only vendors.
Reported cases and investigations
- Amazon — discontinued internal recruiting AI after reports it penalized résumés mentioning "women's" colleges
- HireVue & others — EEOC conciliation and litigation over disability and age discrimination theories
- Netherlands — SYRI welfare profiling case led to government suspension of AI risk scoring
- EU AI Act — employment AI often classified high-risk with conformity and monitoring duties
What deployers must document
Impact assessments, training data representativeness, disparate impact test results, human review pathways, and candidate notification logs. Colorado's AI Act similarly requires deployer impact assessments for high-risk consequential decisions including employment.
Governance workflow
Classify all HR-facing agents as high-risk at intake. Regal AI maps to anti-discrimination controls and assigns bias audit tests in the Prove stage. No deploy without GRC authorization and published audit artifacts where legally required. Runtime logs demonstrate which version of screening logic was live when a candidate was evaluated.