The short answer
NYC Local Law 144 bars employers and employment agencies from using an "automated employment decision tool" (AEDT) to screen candidates or employees for jobs in New York City unless three things are true. The tool had an independent bias audit within the past year. A summary of that audit is public. And candidates who live in the city got notice at least 10 business days before the tool was used. An AI interview that scores candidates and drives who moves forward can fall inside that definition. The city began enforcing the law on July 5, 2023, and the employer, not the vendor, carries the duty.
This is not legal advice. It is a plain-English summary built only from the law's text, the city's rules and the city's own FAQ, all linked at the end. Talk to employment counsel about your specific setup.
Key dates for NYC Local Law 144
We did not find any amendment to the law's requirements on the city's legislation record. Check that record before you rely on this page, because the Comptroller's audit recommended changes to how the city enforces it.
Does your AI interview count as an AEDT?
The law defines an AEDT as a computational process, derived from machine learning, statistical modeling, data analytics or artificial intelligence, that issues a "simplified output" (a score, classification or recommendation) used to "substantially assist or replace" discretionary decision making for employment decisions. Spreadsheets, databases, junk-mail filters and similar tools are carved out.
The city's rules give "substantially assist or replace" a narrow, testable meaning. A tool meets it if you do any of these:
- rely solely on its score, tag, classification or ranking, with no other factors;
- use its output as one criterion, but weight it more than any other criterion; or
- use its output to overrule conclusions from other factors, including a human's judgment.
Two more points from the DCWP FAQ matter for AI interviews. First, "employment decision" includes screening at any stage, not just the final offer. Second, the law covers people who applied for a specific job (or employees up for promotion). Scanning a resume bank, outreach, or inviting people to apply is outside the bias audit and notice duties.
So a practical test: if your recruiters advance whoever clears an AI interview score, or the score outweighs everything else in the packet, treat the tool as an AEDT. If the AI interview produces a transcript and evidence that a recruiter weighs alongside other factors, the analysis is closer, and that is where counsel earns their fee.
When the law applies "in the city"
The FAQ says the law applies when the job's location is an office in New York City at least part time, when the job is fully remote but tied to a New York City office, or when an employment agency using the tool is located in the city. If it applies, the bias audit must come before use, and candidates who are New York City residents must get notice.
What the bias audit must show
A bias audit is "an impartial evaluation by an independent auditor." At a minimum it calculates selection rates or scoring rates, and impact ratios, across sex categories, race/ethnicity categories and intersectional categories (for example, Hispanic women). The rules define the math:
- Selection rate: the share of people in a category who move forward or get a classification. The rules' example: 100 Hispanic women apply, 40 are picked for an interview, so the rate is 40%.
- Scoring rate: for tools that output a score, the share of people in a category who score above the sample's median.
- Impact ratio: a category's rate divided by the rate of the most selected (or highest scoring) category.
Here's a worked example. Say an AI interview advances 60 of 150 men (40%) and 45 of 150 women (30%). The impact ratio for women is 30 divided by 40, or 0.75. The city's rules say these calculations are consistent with the federal Uniform Guidelines (29 CFR 1607.4), where the EEOC's Q&A describes a ratio under four-fifths (0.80) as generally regarded as a sign of adverse impact. Local Law 144 itself does not set a pass mark.
Other audit rules from the FAQ:
- Every year. You can rely on an audit for one year from the date it was done.
- Historical data first. The audit uses data from real use of the tool. Test data is allowed only when there isn't enough historical data, and the summary must say why.
- Pooled data is allowed. An audit can combine several employers' data for the same tool. You can rely on it if you contributed your data, or if this is your first time using the tool.
- No guessing demographics. Imputed or inferred race or sex can't be used.
- Small categories. A category under 2% of the data can be left out of the impact ratio math, with the auditor's justification and that category's counts still reported.
- Results don't force a stop. The law requires the audit but no specific action based on results. Federal, state and city anti-discrimination laws still apply.
Our separate guide to AI bias audits covers how to read an audit summary and what to ask a vendor.
Who counts as an independent auditor
DCWP keeps no approved list. The auditor must exercise "objective and impartial judgment." The FAQ says an auditor is not independent if they work for the employer or the tool's vendor, were involved in using, building or distributing the tool, or hold a direct or material indirect financial interest in either. A vendor may hire an independent auditor to audit its own tool, and many do. But the FAQ is clear that the vendor is not the party responsible: employers and employment agencies must make sure an audit exists before they use the tool.
What you must publish and tell candidates
On the employment section of your website (or through a link to it), publish the date of the most recent audit, the source and explanation of the data, the number of people in an "unknown" category, the number of applicants, the selection or scoring rates, and the impact ratios for all categories. Also post the tool's distribution date, meaning the date you started using it.
The notice to candidates who live in New York City must say you use an AEDT, list the job qualifications or characteristics it assesses, and explain how to request a reasonable accommodation under other laws. Give it at least 10 business days before use, in the job posting, by mail or email, or on the employment section of your website. The law adds that, if not on your website, information about the type and source of data the tool collects and your data retention policy must be provided within 30 days of a candidate's written request.
A simple rollout checklist:
- Decide, in writing, whether the AI interview is an AEDT under the three-part rule test above.
- Get the vendor's current bias audit and check the date, the auditor, and whether it used historical data.
- Post the audit summary and distribution date on your careers page.
- Add the notice to postings or your careers page, then wait 10 business days.
- Put a calendar reminder for the next audit before the one-year mark.
Penalties and enforcement in practice
The law sets a civil penalty of up to $500 for a first violation (and each extra violation that same day), and $500 to $1,500 for each later violation. Each day a tool is used in violation is a separate violation, and each missing candidate notice counts separately. DCWP enforces the audit and notice duties. Discrimination claims go to the NYC Commission on Human Rights.
The State Comptroller's December 2, 2025 audit found DCWP received two AEDT complaints between July 2023 and June 2025. DCWP reviewed 32 companies and found one case of non-compliance, while the Comptroller's auditors reviewing the same companies found at least 17 instances of potential non-compliance. The Comptroller recommended that DCWP look for violations proactively instead of relying mainly on complaints. Low past enforcement is not a reason to skip the steps.
For a vendor example, NTRVSTA has a published NYC Local Law 144 bias audit by the independent auditor Asenion, finalized October 20, 2026, on our compliance page. If you are comparing tools on this point, see our list of AI hiring tools with Local Law 144 bias audits. Hiring outside New York too? Our EU AI Act hiring guide covers Europe, and our AI resume screening guide explains where screening tools fit.
Questions
Does Local Law 144 apply to AI interviews?
It can. If the AI interview issues a score or recommendation that you rely on alone, weight above other criteria, or use to overrule human judgment when screening New York City candidates, it meets the city's definition of an AEDT, and the audit and notice duties apply.
Can we rely on our vendor's bias audit?
Often, yes. An audit can pool several employers' data for the same tool, and you can rely on it if you contributed your own data or are using the tool for the first time. You still own compliance, so check that the audit is under a year old and that the auditor is independent.
What happens if the audit shows a low impact ratio?
Local Law 144 requires the audit and the public summary but doesn't require a specific action based on the results. Anti-discrimination laws still apply, so a low ratio is a signal to investigate the tool, the job criteria and your process with counsel.
Does the law cover remote jobs?
The city's FAQ says it covers fully remote jobs when the location associated with the job is an office in New York City. Notice is owed to candidates who are New York City residents.
Want to see what an audited, adaptive AI interview looks like from the candidate's side? Take a sample interview, or book a 20-minute demo and bring your compliance questions.
Sources
- NYC Council: Local Law 144 of 2021 (Int. No. 1894-A), text and history
- NYC DCWP: Automated Employment Decision Tools
- NYC DCWP: AEDT Frequently Asked Questions (June 2023)
- NYC Rules: DCWP final rule on AEDTs (effective July 5, 2023)
- NYC Rules: DCWP Notice of Adoption (definitions and examples)
- EEOC: Questions and Answers on the Uniform Guidelines (four-fifths rule)
- NY State Comptroller: Enforcement of Local Law 144 (December 2, 2025)
- Our Local Law 144 bias audit summary
