By Amanda D. Allen and Bre Timko
BLOG OVERVIEW: Despite the Trump Administration’s Executive Order 14281 directing federal agencies to deprioritize disparate impact enforcement and the EEOC’s decision to close disparate-impact-only investigations, four states—California, New York, New Jersey, and Illinois—have recently enacted laws that codify disparate impact liability at the state level, with some formally adopting the Uniform Guidelines on Employee Selection Procedures (UGESP).
In April 2025, the Trump Administration issued Executive Order 14281, “Restoring Equality of Opportunity and Meritocracy,” directing federal agencies to deprioritize disparate impact analysis and stating it is U.S. policy to eliminate disparate impact liability “to the maximum degree possible.” The Equal Employment Opportunity Commission (EEOC) subsequently announced it would no longer investigate cases based solely on disparate impact claims, effective September 30, 2025.
Disparate impact liability allows challenges to policies that appear facially neutral but result in disproportionate harm to protected groups even without discriminatory intent. EEOC has historically relied on this principle in areas such as recruitment, hiring, promotions, and terminations. These developments are particularly significant as the use of artificial intelligence (AI) in hiring continues to grow and concerns about algorithmic discrimination increase.
Four states have recently enacted laws that codify disparate impact liability and/or incorporate the Uniform Guidelines on Employee Selection Procedures (UGESP), which has long been the federal framework governing how employers evaluate and validate selection procedures to ensure compliance with nondiscrimination requirements, reinforcing the continued relevance of this theory at the state level despite recent federal efforts to limit its enforcement.
- California (Effective October 1, 2025): The Fair Employment and Housing Act (FEHA) Amendments formally codify disparate impact liability and adopt UGESP, requiring employers to justify practices that produce adverse impact as job-related and consistent with business necessity, and demonstrate that no less discriminatory alternative would achieve the same objective. Employers may defend against discrimination claims by demonstrating proactive anti-bias testing and mitigation efforts.
- New York (Effective December 19, 2025): The New York State Human Rights Law (NYSHRL) Amendment explicitly codifies disparate impact liability, establishing that “an unlawful discriminatory practice may be established by a practice’s discriminatory effect, even if such practice was not motivated by a discriminatory intent.”
- New Jersey (Effective January 1, 2026): The Rules Pertaining to Disparate Impact Discrimination under New Jersey’s Law Against Discrimination also formally codifies disparate impact liability and incorporates UGESP, establishing a burden-shifting framework requiring complainants to demonstrate disparate impact and employers to justify practices with empirical evidence supporting substantial, legitimate interests (i.e., validation evidence). Plaintiffs may prevail by identifying a less discriminatory alternative, even if it is not equally effective.
- Illinois (Effective January 1, 2026): Amendments to the Illinois Human Rights Act explicitly prohibit the use of AI that results in discrimination based on protected characteristics, even if unintentional. While the statute uses effects-based language (“has the effect of”) rather than explicitly naming disparate impact, the Illinois Human Rights Commission interprets and enforces it as such, confirming that disparate impact claims will be investigated under this law.
Recommendations for Employers
So how should employers respond to this increasingly complex contrast between federal and state regulatory landscapes?
- Audit All Selection Procedures Used in Employment Decisions
Employers should conduct a thorough audit of all selection procedures used in recruitment, hiring, promotions, and terminations, including AI and algorithmic tools. Even well-intentioned, facially neutral tools can produce discriminatory outcomes for protected groups. Proactive bias testing, now recognized as an affirmative defense in California, can reduce legal exposure and demonstrate good-faith compliance efforts. - Familiarize Yourself with UGESP
California and New Jersey have formally incorporated UGESP into their state frameworks, and other states may follow. Under UGESP, any practice producing adverse impact must be validated as job-related and consistent with business necessity, which is supported by empirical evidence, or replaced with a less discriminatory alternative. Employers should treat UGESP as the baseline standard for justifying and defending their selection procedures. - Do Not Assume Federal Inaction Equals Reduced Risk
EEOC’s decision to stop investigating disparate impact-only claims may create a false sense of security. State agencies in California, New York, New Jersey, and Illinois retain full authority to enforce disparate impact claims. Employers operating in these states should maintain or strengthen their compliance programs regardless of the federal posture. - Monitor Emerging State Legislation
Additional states may adopt similar language codifying disparate impact liability, particularly as AI use in hiring grows. Employers, especially those operating across multiple states, should track legislative and regulatory developments closely and build flexibility into their HR policies and technology procurement processes to adapt as the legal landscape continues to evolve.
DCI will continue to monitor developments regarding the regulation of AI tools in employment. Sign up to receive alerts and information about upcoming webinars here.
