AI in Hiring & Employment Decisions

No federal law, but federal liability is real

Congress has not enacted any AI-specific hiring statute. But Title VII, the ADA, and the ADEA all apply to AI-driven employment decisions, and the EEOC has made clear employers cannot blame their vendors. The landmark Mobley v. Workday class action — alleging a screening vendor's AI discriminated by race, age, and disability — is still active in the Northern District of California and will shape how courts allocate liability between employers and AI tool providers.

States are filling the void — and the federal moratorium is dead

The U.S. Senate voted 99-1 in July 2025 to strip a 10-year moratorium on state AI laws from the One Big Beautiful Bill Act. Six jurisdictions now have enacted laws in force or effective by mid-2026: NYC, Illinois (two laws), Maryland, Colorado, Texas, New Jersey, and California. 1,561 additional AI bills have been introduced in 45 states as of March 2026. The patchwork is accelerating, not consolidating.

The Trump EO on AI preemption is a signal, not a shield

EO 14365 (December 2025) directed the DOJ to form an AI Litigation Task Force to challenge state AI laws and directed Commerce to identify "onerous" provisions. As of May 2026, no state AI hiring law has been enjoined. Employers should plan for compliance with state and local laws — EO 14365 does not pause enforcement obligations while litigation is pending.

The laws diverge sharply in what they require

NYC requires annual independent bias audits. Colorado requires impact assessments and a risk management program. Illinois mandates notice and prohibits discriminatory outcomes. New Jersey codifies disparate impact liability with AI tools explicitly in scope. Texas only prohibits intentional discrimination with a 60-day cure period. California uses two parallel tracks (FEHA + CPPA). A single compliance posture does not work across all jurisdictions — employers need a jurisdiction-by-jurisdiction inventory.