evidoria

← Back to browse

Good practice

New York City Local Law 144 — AI Hiring Bias Audit Mandate

United States of America · New York City · See the United States of America profile

NYC's Local Law 144 (2021), the first law to require bias audits of AI hiring tools by sex and race, took effect July 2023. A 2024 study of 391 employers found only 4.6% posted audit reports; a 2025 state audit called enforcement 'ineffective'.

New York City Local Law 144 — AI Hiring Bias Audit Mandate New York City Local Law 144 — AI Hiring Bias Audit Mandate

Details

Promoter
NYC Department of Consumer and Worker Protection (DCWP)
Period
Enacted December 2021; enforced July 2023–present
Keywords
algorithmic accountability, employment law, AI governance, gender & race equity

Description

New York City's Local Law 144, enacted in December 2021 and enforced from July 2023, requires any employer using an Automated Employment Decision Tool (AEDT) — algorithmic software used to screen, score or rank candidates — to commission an independent annual bias audit. The audit must calculate selection and scoring rates broken down by sex, race/ethnicity and their intersections, publish a summary on the employer's website, and notify candidates that an AEDT is in use with the right to request an alternative process. It was the first law globally to mandate algorithmic bias audits for hiring.

Independent evidence of the law's real-world performance is mixed to poor. A peer-reviewed FAccT 2024 study ('Null Compliance'), in which 155 student investigators tracked 391 New York employers, found only 18 (about 4.6%) had posted a bias audit report and just 13 (about 3.3%) had posted the required transparency notice. The researchers coined the term 'null compliance' because the law gives employers wide discretion over whether their tool is even in scope, making it impossible to tell missing disclosures apart from lawful exemptions.

A New York State Comptroller audit released in December 2025 corroborated this picture at the regulator level: it found 75% of test calls to the city's 311 hotline about AEDT complaints were misrouted and never reached DCWP, that DCWP's own review of 32 posted bias audits found only one compliance issue where the Comptroller's auditors found at least 17, and that DCWP received only two formal AEDT complaints in the two-year audit period despite penalties of $500–$1,500 per violation being available. The Comptroller concluded DCWP's enforcement was 'ineffective.'

The law remains significant as a first-mover model that explicitly frames algorithmic hiring bias as a sex- and race-equity issue, and it has shaped subsequent AI-hiring proposals elsewhere. But the documented evidence to date is a cautionary case: transparency-only, employer-self-scoped regulation has not yet produced verifiable compliance or measurable bias reduction.

Read the full analysis: https://www.nyc.gov/site/dca/about/automated-employment-decision-tools.page

Implementation

Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful