Archbright Blog

California's New AI Employment Regulations: What Employers Need to Know

Written by Ifeoma Enenmoh | Aug 14, 2026, 4:24:53 PM

Artificial intelligence (AI) is rapidly becoming part of the modern workplace. From screening resumes to evaluating applicants, employers are increasingly relying on technology to streamline employment decisions. At the same time, new state regulations, such as those in California, impose limitations on AI’s authority. They serve as an important reminder that while AI can support decision-making, employers remain responsible for ensuring those decisions comply with anti-discrimination laws.

Archbright members and mozzo subscribers can watch a video briefing on these updates in mozzo.

Watch the video →

What's Changing?

Effective October 1, 2025, regulations under the California Fair Employment and Housing Act (FEHA) clarify how existing discrimination protections apply to employers' use of AI and other automated decision systems (ADS). The regulations do not prohibit the use of AI. Instead, they establish expectations for how employers evaluate, implement, and monitor these tools. They also focus not only on intentional discrimination but also on practices that may have a disparate impact on applicants or employees based on protected characteristics.

What Is an Automated Decision System?

The regulations broadly define an ADS as a computational process that makes or assists with employment-related decisions. This may include technologies that use AI, machine learning, algorithms, statistical modeling, or similar methods. Employers can use these systems to:

  • Screen or rank job applicants
  • Review resumes for specific qualifications
  • Assess candidates through online tests or games
  • Analyze video or audio interviews
  • Recommend employees for promotions, transfers, or training opportunities
  • Target recruiting advertisements for specific audiences

While many covered tools incorporate AI, the regulations apply more broadly to ADS and are not limited solely to generative AI technologies. Common office software, such as word processing or spreadsheet applications, is generally not covered by the regulations.  

Takeaways For Employers

AI Doesn't Shift Liability Away from Employers


Employers remain accountable for employment decisions, even when AI helps inform or make these decisions. If an automated system disproportionately excludes or disadvantages applicants or employees based on a protected characteristic, an employer may still face liability under FEHA. Reliance on a third-party vendor is also not a defense if a tool produces discriminatory outcomes. Employers should understand how AI tools work, what decisions they influence, and whether they have been evaluated for bias before implementation.

Reasonable Accommodations Still Matter


Employers must also ensure that AI-driven processes align with existing obligations to provide reasonable accommodations. For example, an applicant who cannot complete an automated assessment because of a disability or who requires an accommodation based on a sincerely held religious belief may need an alternative process. Technology should not create unnecessary barriers for qualified applicants or employees.

Recordkeeping Requirements Have Expanded


The regulations also broaden record retention expectations. Employers using ADS should retain records related to these systems, including scoring outputs, datasets, audit information, and other relevant documentation, for at least four years. Maintaining these records can be critical when responding to agency inquiries, legal claims, or questions about how employment decisions were made.

Assessing AI Vendors Is Critical


Many employers obtain AI-enabled HR technology from third-party vendors. Even when vendors market products as compliant or bias-tested, employers should conduct their own due diligence. Employers should ask:

  • Has the tool been evaluated for bias or disparate impact?
  • How frequently is the system tested?
  • What data was used to develop and validate the technology?
  • Will the vendor provide documentation regarding testing and methodology?
  • How are system updates and modifications evaluated?

The answers to these questions can help employers better understand potential risks and make more informed decisions about implementation.

Next Steps For Employers

Organizations that use AI, even in limited ways, should take this opportunity to evaluate current practices.

  • Identify where AI or ADS are being used throughout the employment lifecycle.
  • Review vendor relationships and request information regarding testing, validation, and data governance.
  • Consider periodic bias testing or audits to assess whether AI-assisted employment processes disproportionately affect individuals in protected groups.
  • Maintain meaningful human oversight before employment decisions are finalized.
  • Update record retention practices to include AI-related documentation.
  • Train HR professionals, recruiters, and managers on the legal considerations surrounding AI-assisted decisions.
  • Review accommodation procedures to ensure applicants and employees can request alternatives when needed.

The Bottom Line

California's new regulations reinforce a straightforward principle—employers are responsible for the employment decisions they make, regardless of whether those decisions are assisted by artificial intelligence. Organizations that take proactive steps to understand, monitor, and document their use of AI will be better positioned to reduce legal risk while benefiting from emerging technology.

Need Assistance?

If you have questions about how these new regulations will impact your organization, get in touch. Archbright members and mozzo subscribers can reach out to our advisor team for support via the HR Advisor Hotline at at 888.622.4402 ext. 2 or Advisor Chat in mozzo. Not yet a member? Connect with info@archbright.com to learn more about the benefits of membership