New York Codifies Disparate Impact Liability Under The State Human Rights Law

Key Highlights:

• A recent amendment expressly codifies disparate impact liability under the New York State Human Rights Law (NYSHRL) for employment discrimination claims. This comes as the U.S. Equal Employment Opportunity Commission has backed away from disparate impact theories in enforcing federal employment discrimination statutes.
• The increasing use of Artificial Intelligence (AI) tools in personnel processes and decision-making has the potential to raise disparate impact issues to the extent that AI processes have varying effects on specific groups.
• New York employers may face increased potential exposure from neutral employment practices, underscoring the importance of proactive review and documentation.

New York Governor Kathy Hochul signed Senate Bill S8338 on Dec. 19, 2025, which codifies that a facially neutral employment practice may violate the New York State Human Rights Law (NYSHRL) based on its discriminatory effects, even absent discriminatory intent. While the amendment largely clarifies existing law, it comes at a time when federal enforcement of disparate impact theories has become less certain as the U.S. Equal Employment Opportunity Commission has taken a more restrained approach to pursuing disparate impact claims under federal employment discrimination statutes. Against that backdrop, the amendment underscores the continuing importance of state-law compliance and employer attention to outcome-based employment practices, as well as the purportedly neutral decisions of their AI tools.

What the Amendment Does

The legislation adds a new subdivision to New York's Executive Law § 296, providing that, in NYSHRL employment discrimination cases, an unlawful discriminatory practice may be established where an employer uses a policy or practice that actually or predictably results in a disparate impact based on a protected characteristic.

The statute makes clear that proof of discriminatory motive is not required. After the employee demonstrates that a particular employment practice causes, or predictably will cause, a disparate impact on a protected class, the employer then bears the burden to establish that the practice is job-related for the position in question and consistent with business necessity. Even if that showing is made, an employee may still prevail by showing the employer's business necessity could be satisfied by a less discriminatory alternative.

The statute also requires that an employer's justification be supported by evidence and not based on hypothetical or speculative considerations, reinforcing the need for objective validation and documentation of employment criteria.

Although disparate impact liability is not a new concept, the amendment injects ambiguity into the analysis by prohibiting policies and practices that "actually or predictably" yield disparate results. This raises the specter of challenges to practices that do not "actually" cause a disparate impact but can be argued to "predictably" do so. Given the litigation climate in New York, this additional language creates another reason for employers to be intentional in assessing the effect, or event "predicted" effect,of personnel practices and policies.

Why This Matters Now, Especially as AI Gains Ground in Employment Practices

By codifying disparate impact liability, New York has increased scrutiny of ostensibly neutral employment practices — such as hiring criteria, screening tools, promotion standards and compensation structures — that may produce statistically significant disparities.

AI tools are often adopted to promote efficiency and consistency and typically would not be viewed as intentionally discriminatory. However, these tools present disparate impact risks to the extent that the data inputs, models or selection criteria underlying those tools have varying effects on specific groups. For example, even in the relatively early phases of AI's adoption, there have been claims in litigation that an employer's use of AI training datasets disproportionately composed of one protected group (for example, males) results in an adverse disparate impact to members of other groups (for example, females).

The beefed-up disparate impact liability under NYSHRL, combined with New York City's 2023 regulations on AI use in personnel processes, guide in favor of an intentional approach by employers in using these tools for employment decisions.

Looking Ahead

This amendment applies to employment discrimination occurring on or after its effective date of Dec. 19, 2025, making proactive compliance efforts particularly important. Employers should consider reviewing key employment practices to assess disparate impact risk, ensure that job-related criteria are well supported, and evaluate whether alternative approaches could achieve business objectives with less discriminatory effect. If you have questions about how this amendment may affect your organization, or would like assistance evaluating existing policies and practices, contact your Polsinelli Labor & Employment attorney.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.