NYC Local Law 144: The Automated Employment Decision Tools Compliance Guide
New York City was the first U.S. jurisdiction to require an independent bias audit of AI hiring tools. It is still the only one. Here is what the law covers, what the audit must show, and what employers and vendors each need to do.
Who the law applies to
Local Law 144 applies to employers and employment agencies that use an automated employment decision tool (AEDT) to screen candidates for hiring, or employees for promotion, for jobs in New York City. That includes remote roles associated with a New York City office, and it applies wherever the employer is headquartered.
An AEDT is a computational process, derived from machine learning, statistical modeling, data analytics or AI, that issues a simplified output such as a score, classification or recommendation, and that is used to substantially assist or replace discretionary decision-making. Resume screeners that rank candidates, interview tools that score answers and assessments that recommend who advances are typical examples.
The three core obligations
- An independent bias audit completed no more than one year before the tool is used, and repeated at least annually while it stays in use.
- A public summary of the most recent audit posted on the employment section of the employer’s website before use, including the audit date and the tool’s distribution date.
- Notice to candidates and employees who reside in New York City, at least 10 business days before the tool is used, saying an AEDT will be used, which job qualifications and characteristics it assesses, and how to request an alternative selection process or accommodation if one is available.
What the bias audit must calculate
The audit is a disparate-impact analysis. For each category it reports a selection rate (or, for tools that output a score, a scoring rate: the share of people scoring above the median) and an impact ratio, which compares each category’s rate to the rate of the most-selected category.
- Sex categories
- Race and ethnicity categories
- Intersectional categories of sex combined with race and ethnicity
- The number of individuals the tool assessed who fall into an unknown category
An auditor may exclude a category that represents less than 2% of the data, but must say so and report that category’s count. The audit should use historical data from the employer’s own use of the tool where enough exists. Where it does not, test data can be used, with an explanation of why.
What “independent” means
The auditor cannot have been involved in using, developing or distributing the tool, cannot have an employment relationship with the employer or vendor during the audit, and cannot have a direct financial interest in either. A vendor’s in-house fairness report is useful evidence, but it is not an independent audit.
Employers and vendors: who does what
The obligation sits with the employer or employment agency, not the vendor. In practice, vendors often commission an audit across their customers’ data so that each employer can rely on it. That only works if the employer’s own data is included or the employer has none yet. Employers should check that a vendor-supplied audit actually covers their use of the tool.
Penalties and enforcement
Violations carry civil penalties of up to $500 for a first violation and between $500 and $1,500 for each later one, and each day a tool is used without compliance can count as a separate violation. In December 2025 the New York State Comptroller found the city’s enforcement had been weak, and the Department of Consumer and Worker Protection has since committed to tighter enforcement. Expect more scrutiny, not less.
A practical checklist
- List every tool in your hiring and promotion stack that scores, ranks or recommends candidates.
- Decide for each whether it substantially assists or replaces human judgment.
- Commission an independent audit at least annually, using your own historical data where you have it.
- Publish the summary and keep the notice live 10 business days before use.
- Keep the audit data and methodology on file so you can answer a regulator or plaintiff.
Map your hiring AI against the laws that apply to you
Whether you build HR AI or deploy it, Fairly AI tests it, monitors the rules where you operate and keeps the evidence.
Schedule a discovery call