AI governance in US employment & hr.

Employers are liable for discriminatory AI hiring outcomes regardless of whether a vendor designed the tool. Disparate impact analysis applies, if AI produces significantly lower selection rates for a protected group, this creates legal exposure.

NYC employers using Automated Employment Decision Tools must conduct annual independent bias audits, publish results publicly, and notify candidates 10 business days before AEDT use. Non-compliance: $375–$1,500 per violation.

Never took effect: SB 24-205 was repealed and replaced in May 2026 by SB 26-189, a narrower disclosure law effective 1 January 2027. From that date Colorado employers using AI in employment decisions face ADMT disclosure duties.

Employers must notify applicants before AI video interview analysis, explain how the AI works, obtain consent, and delete videos within 30 days of request. Civil penalties apply to violations.

Workers discussing and taking collective action on workplace AI use are protected from retaliation under the NLRA. Mandatory arbitration clauses cannot prevent workers from filing unfair labor practice charges related to AI-driven management decisions.

Connecticut and New York require advance written notice of electronic monitoring including AI productivity tools. Delaware and several other states have similar requirements.