Quick Hit
The California Civil Rights Board has final approval of comprehensive regulations governing the use of AI and “automated decision systems” aimed at preventing discrimination. These regulations require employers not to use an “automated decision system” that discriminates against applicants or employees based on protected characteristics under California’s differentiation laws. California will join other states in implementing AI regulations for employment decisions, continuing to consider additional legislation to manage the impact of emerging technologies in the workplace. The new California regulations are expected to come into effect on October 1, 2025.
On June 27, 2025, the California Administrative Law Office submitted a notice of approval for the latest amended text of the proposed regulations that the California Office of Civil Rights (CRD) advanced in March 2025. Public comment letter.
The final regulations make it clear that it is illegal for employers to use “automated decision systems” or selection criteria that discriminate against applicants or employees on standards protected by California’s Fair Employment and Housing Act (FEHA) and other California anti-discrimination laws.
basis
In its approval statement, the Civil Rights Council states that “automated decision systems” relying on algorithms or AI are “increasingly used in employment settings to promote a wide range of decisions regarding job seekers or employment, employment and promotion.” These technologies “have countless benefits, but they can exacerbate existing biases and contribute to discriminatory outcomes,” says the Civil Rights Council.
“These new regulations on artificial intelligence in the workplace aim to help states anti-discrimination protections stay at their pace,” CRD director Kevin Kisch said in a statement. “I commend the Civil Rights Council for their commitment to protecting the rights of all Californians.”
Important Terminology
The final rule amends existing regulations to define key terms related to AI.
“Automated decision systems” are defined as “a computational processes that make decisions about employment benefits or facilitate human decisions about employment benefits,” including processes that may be derived and/or used from artificial intelligence, machine learning, algorithms, statistics, and/or other data processing technologies. “Agents” – The final regulations consider the employer’s agent to be “employer” under the FEHA regulations. Specifically, regulations define “agents” as “including when performed in part or in part through the use of automated decision-making systems to act directly or indirectly on behalf of an employer to exercise functions traditionally exercised by an employer or other FEHA regulated activities.” (Embodiment added.) “Automatic Decision System Data” – Regulations cover “(a) NY data used to develop or customize automated decision systems used by certain employers or other target companies.” “Artificial Intelligence” – Regulations define AI as “a machine-based system that infers how output is generated from the input it receives” and can include “prediction, content, recommendation, or decision.” “Machine Learning” – The term is defined as “the ability of a computer to use and learn from an analysis of a data or experience and to automatically apply this learning in future calculations or tasks.”
Illegal Selection Criteria
The rules further clarify that California law bans discriminatory employment tools in automated decision systems or AI tools. According to the regulations, it is “declared that it is illegal for FEHA to use an automated decision system or selection criteria (including eligibility criteria, employment tests, or proxies) that discriminate against the applicant or employee or class of employees in standards protected by FEHA “to be illegal for an employer or other eligible companies.”
Next Steps
Regulations allow California to participate in more and more states and jurisdictions, including Colorado, Illinois, and New York City, including those used to make employment-related decisions to implement laws or regulations that enact AI and similar technologies. At the same time, California continues to propose regulations to manage overlapping parts of the law and AI. In particular, the bill, called the “Roboboss-free Act,” requires employers to provide human supervision over the use of AI.
State law comes as President Donald Trump’s administration tried to remove legal restrictions on AI to promote technological development in the United States. The administration’s spending bill included a 10-year moratorium that banned states from enacting or enforcing AI laws and regulations. However, U.S. Senate lawmakers revised the proposal and were eventually removed from the bill passed by the Senate.
New California regulations are expected to come into effect on October 1, 2025. California employers may want to review new regulations and consider how they affect operational and employment decision-making policies and practices, including hiring, employment, promotions, and disciplinary decisions.
