NYC AI enforcement in 2026: why employers remain responsible for AI hiring decisions.

What NYC announced on October 1, 2026

The agencies' joint statement addresses AI, algorithms and automated systems used across employment, housing, consumer services and transportation. In the workplace, the agencies identify uses involving candidate selection, hiring, pay, scheduling, timekeeping, accommodations, evaluation, discipline and termination. Their enforcement position is that technology does not displace the obligations imposed by the laws already in force.

Legal-status distinction: this is an official agency enforcement policy statement, not a newly enacted Local Law, amended LL144 regulation or finding that a particular employer broke the law. The joint document expressly says that it announces no new legal standards or requirements.

What does this mean for employers using AI to screen candidates?

A company using algorithmic screening must ask two different questions. First, does the system qualify as an automated employment decision tool (AEDT) covered by NYC Local Law 144? Second, independently of LL144's technical definition, does the company's use of technology comply with other applicable employment and anti-discrimination protections?

DCWP's LL144 guidance covers tools that substantially assist or replace discretionary decision-making for covered hiring or promotion screening in the city. The October statement is broader: its application of existing anti-discrimination and worker-protection laws is not limited to tools that meet LL144's AEDT definition. Conversely, the statement does not make every workplace algorithm an AEDT under LL144.

Who is responsible: employer or software vendor?

For LL144's bias-audit requirement, the employer or employment agency using a covered AEDT remains responsible for ensuring that a qualifying audit has been completed before use. DCWP expressly states that the vendor creating the AEDT is not the party responsible for that LL144 requirement. A vendor may arrange for an independent audit, but the user of the tool must verify that the statutory and regulatory prerequisites for its own use are satisfied.

Vendor contracts, technical assurances or a vendor-hosted audit summary may be helpful evidence; they are not automatic exemptions from LL144. The same principle should not be overstated: responsibility under a particular discrimination or consumer-protection claim depends on the governing law and facts, and the joint statement does not decide private liability in an individual dispute.

Does an LL144 bias audit prove the AI system is non-discriminatory?

No. LL144 requires an independent bias audit with specified statistical measures, including selection or scoring rates and impact ratios for demographic categories defined in the rules. DCWP's FAQ expressly notes that LL144 itself does not impose a specific corrective action solely because an audit indicates disparate impact. It also warns that employers and employment agencies must comply with applicable federal, state and city anti-discrimination laws, which may require further action depending on the facts.

Accordingly, an audit can document a required assessment, but it should not be marketed as a blanket legal certification, a finding of no discrimination or immunity from claims. See our bias-audit process and synthetic example of a public Summary of Results for the different types of evidence involved.

LL144 obligations that have not changed

If an employer or employment agency uses a covered AEDT in NYC, the underlying LL144 requirements continue to include:

• A recent independent bias audit: the tool must have been audited no more than one year before covered use.
• Public disclosure: the latest required Summary of Results and the AEDT's distribution date must be publicly available before use.
• Candidate and employee notices: the required disclosures must be supplied to covered NYC-resident candidates or employees at least 10 business days before use, including information about the assessed qualifications and how to request an alternative process or accommodation under the statutory and regulatory framework.
• Scope and data review: determine whether the system and proposed use meet the AEDT definition, and whether the data and audit methodology satisfy applicable DCWP rules.

These are summaries, not substitutes for the statutory text, DCWP FAQ and full LL144 guide. Some data and notice details depend on the factual deployment context.

Practical AI hiring governance: evidence employers can preserve

The following are risk-management recommendations, not newly mandated steps in the October 1 statement:

1. Map each AI-assisted hiring and promotion workflow, including the tool, version, decision point, human involvement and relevant NYC nexus.
2. Request vendor documentation describing outputs, material functionality, data origin, model changes and limitations, then independently assess whether LL144 applies.
3. Confirm the independent auditor's eligibility, audit date, permitted data basis, required calculations and public-summary content; do not assume a generic vendor report meets LL144.
4. Check that required notices and the public Summary of Results are available before covered use, and keep dated evidence of those disclosures.
5. Escalate potential discriminatory outcomes or adverse-impact signals for a separate employment-law review, rather than treating an LL144 audit alone as a complete discrimination assessment.

For organizations still defining data availability and audit scope, start with the data-readiness checker or audit-readiness review.

Frequently asked questions

Can an employer rely on an AI vendor's bias audit?

Potentially, subject to LL144's rules on independent auditing and appropriate historical or test data. But DCWP places the compliance obligation on employers and employment agencies that use a covered AEDT, not automatically on the software vendor.

Is the October 2026 statement a new NYC AI hiring law?

No. It is an enforcement policy statement about existing DCWP, CCHR and TLC legal protections. It does not amend LL144 or introduce a new independent bias-audit regime.

Does passing an LL144 bias audit mean the hiring system is legally safe?

No. LL144's audit and disclosure requirements do not replace separate anti-discrimination obligations. A statistical audit may identify issues requiring examination under other laws, and no audit removes fact-specific legal risk.

Related AUDITLL144 resources

LL144 statutory compliance guide · HR software vendor bias audits · Proposed NYC AI bills: separate legal status · Enforcement and oversight tracker

Request an independent LL144 audit scope review

Prepared by AUDITLL144 Research Team, a research and publishing function of Lexara Advisory LLC. General information, not legal advice; no legal representation or outcome is promised.