What you'll learn
- What Local Law 144 Actually Requires
- Why Vendor Assurances Alone Don't Satisfy the Law
- The Candidate Notice and Alternative-Process Requirement
- Penalties and the Practical Risk Beyond the Statute Itself
- Building a Compliant AI Hiring Process
NYC Local Law 144 has become the reference point for AI hiring tool regulation in the US, but the specific compliance requirements — an independent bias audit, public disclosure, and a real candidate notice-and-alternative-process mechanism — are frequently misunderstood, particularly the independence standard that disqualifies a vendor's own internal fairness testing from counting as compliance. This guide covers exactly what the law requires, why vendor-supplied assurances alone don't satisfy it, the notice and alternative-process obligation most employers under-implement, the actual penalty exposure, and how to build an AI hiring process that holds up to genuine scrutiny.
What Local Law 144 Actually Requires
Quick answer
New York City's Local Law 144, effective July 5, 2023, applies to any employer or employment agency using an automated employment decision tool (AEDT) to screen candidates for employment or promotion within New York City. The law defines an AEDT broadly: any computational process derived from machine learning, statistical modeling, data analytics, or artificial intelligence that issues a simplified output — a score, classification, or recommendation — that is used to substantially assist or replace discretionary decision-making in hiring or promotion.
Three distinct obligations apply to a covered AEDT. First, the employer must obtain an independent bias audit of the tool conducted within one year prior to its use, evaluating selection or scoring rates across sex, race, and ethnicity categories using an impact ratio methodology. Second, a summary of the audit results must be made publicly available — typically posted on the employer's own website — before the tool is used on any NYC candidate. Third, candidates must receive advance notice that an AEDT will be used in their evaluation, along with instructions for requesting an alternative evaluation process or accommodation if they prefer not to be assessed by the tool.
The law applies regardless of where the employer is headquartered — what matters is whether the position is located in New York City, or whether the candidate being screened is a NYC resident applying for a role that could be performed there. A national employer using an AI screening tool across its entire hiring pipeline is covered for any NYC-based candidate or NYC-based role, even if the company itself has no NYC office.
Why Vendor Assurances Alone Don't Satisfy the Law
Quick answer
The single most common misunderstanding among employers evaluating AI hiring tools is treating a vendor's internal fairness testing, or a general marketing claim of being 'bias-free' or 'EEOC-compliant,' as equivalent to a Local Law 144 bias audit. It is not. The law specifically requires the audit be conducted by an independent auditor — one without a current employment, financial, or other relationship with the AEDT's developer or the employer that would compromise the audit's objectivity.
This independence requirement exists precisely because a vendor's own internal testing, however rigorous, carries an inherent conflict of interest: the vendor has a direct commercial incentive to produce a favorable result. An employer relying on vendor-supplied fairness documentation without confirming it was produced by a genuinely independent third-party auditor is very likely not in compliance, regardless of how thorough the vendor's own materials appear.
Employers evaluating an AI hiring vendor should specifically request the auditor's identity and confirm the absence of a financial or reporting relationship with the vendor, rather than accepting a general assurance that 'our tool has been audited.' A vendor unwilling or unable to identify the specific independent auditor and provide the actual audit summary — not just a marketing claim referencing one — is a meaningful red flag during procurement.
Local Law 144 doesn't ban AI hiring tools — it requires three specific things before one can be used to screen NYC candidates: an independent bias audit within the prior year, public posting of a results summary, and advance notice to candidates with an alternative-process option, and missing any one of the three is a violation even if the tool itself is genuinely unbiased.
The Candidate Notice and Alternative-Process Requirement
Quick answer
Beyond the audit itself, Local Law 144 requires employers to notify candidates at least ten business days before an AEDT is used to screen them for a position or promotion, unless the notice is provided through a job posting, in which case the timing requirement is satisfied by the posting itself being available at least ten business days before the tool's use. The notice must explain that an AEDT will be used, identify the job qualifications and characteristics the tool assesses, and inform the candidate they may request an alternative selection process or accommodation.
This alternative-process requirement is frequently under-implemented in practice — many employers add the required notice language to a job posting but never actually build a functioning process for a candidate who requests an alternative to the AI tool. A notice that technically discloses the AEDT's use, paired with no genuine mechanism for a candidate to opt into a different evaluation path, satisfies the letter of the disclosure requirement while leaving the substantive alternative-process obligation unmet.
Building this into the hiring workflow means having an actual designated process — a specific point of contact, a defined alternative evaluation method (such as a live human screen in place of the AI tool), and internal training so recruiters know how to handle such a request when it arrives, rather than treating the requirement as satisfied purely by the presence of boilerplate notice language.
Penalties and the Practical Risk Beyond the Statute Itself
Quick answer
The NYC Department of Consumer and Worker Protection (DCWP) enforces Local Law 144, with civil penalties for violations ranging from $500 for a first violation to $1,500 for each subsequent violation, with each day of continued noncompliance and each candidate affected potentially treated as a separate violation — meaning the actual financial exposure for a systemic, ongoing compliance gap can scale well beyond the per-violation figure suggests at first glance.
Beyond the direct DCWP penalty exposure, an employer's use of an unaudited or improperly audited AEDT creates real exposure under separate federal and state anti-discrimination law if the tool's actual selection outcomes turn out to reflect a disparate impact — the EEOC has been explicit that employers bear liability for AI hiring tool outcomes regardless of vendor indemnification language, meaning Local Law 144 compliance and broader disparate-impact risk management are closely linked, not separate concerns.
Given the relative youth of this regulatory area, employers should also expect continued enforcement guidance and potential legislative expansion — several other jurisdictions have proposed or are actively developing similar AEDT-style requirements, meaning a compliance framework built now around Local Law 144's specific standard is a reasonable template to extend as similar laws take effect elsewhere.
Related reading
Building a Compliant AI Hiring Process
Quick answer
Start by confirming whether each AI tool in your hiring stack — interview scoring, resume screening, candidate ranking — actually meets the AEDT definition, since not every hiring technology qualifies, and treating every piece of recruiting software as automatically covered wastes compliance effort on tools that don't need it while potentially missing ones that genuinely do.
For any tool that does qualify, confirm the specific independent auditor's identity and obtain the actual audit summary directly, build the ten-business-day candidate notice into your application flow with genuinely accessible language, and stand up a real, trained process for handling alternative-process requests rather than treating the requirement as satisfied by notice text alone.
For a deeper look at how AI interview platforms are evaluated on compliance readiness more broadly — not just Local Law 144, but the Illinois AI Video Interview Act and emerging state requirements — see our guide on AI video interview platforms and our overview of the best AI interview software in 2026, both of which weigh vendor compliance documentation as a core evaluation criterion, not an afterthought.
The independence requirement for the bias audit is the piece most vendors and employers underestimate — an auditor with a financial or reporting relationship to the tool's vendor cannot produce a compliant audit, which means a vendor's own internal fairness testing, however rigorous, does not satisfy the law on its own.
Frequently asked questions
Common questions about ai recruiting and how InCruiter helps teams solve them.
InCruiter Editorial Team
AI Hiring Research · Interview Intelligence · Enterprise Talent Strategy
The InCruiter editorial team covers AI-driven hiring, interview intelligence, and modern talent acquisition strategy. Our guides draw on platform data from 2,000+ hiring teams, conversations with talent leaders, and published research in industrial-organizational psychology.
