California Civil Rights Council Employment Regulations Regarding Automated-Decision Systems

What it is

The California Civil Rights Council, which makes regulations under the Fair Employment and Housing Act (FEHA), adopted modifications to its employment regulations to address automated-decision systems. The Office of Administrative Law approved them on 27 June 2025, and they took effect on 1 October 2025. They sit in Title 2 of the California Code of Regulations, starting with a new section 11008.1 of definitions.

The regulations do not create a new AI law. They make clear how California’s existing employment discrimination law applies when software makes or shapes employment decisions, and they adjust the rules on recruitment, pre-employment inquiries, medical inquiries, criminal history and record-keeping so that automated systems are expressly covered.

Who it applies to

The regulations apply to employers and other entities covered by FEHA, broadly employers of five or more people in California, along with employment agencies, labour organisations and apprenticeship programmes.

An automated-decision system is a computational process that makes a decision or facilitates human decision-making about an employment benefit. It may use AI, machine learning, algorithms, statistics or other data processing. The regulations list examples.

• computer-based tests, questions, puzzles or games that predict, measure skills or abilities, measure personality, aptitude or cultural fit, or screen and rank applicants
• directing job adverts or recruiting material to targeted groups
• screening CVs for particular terms or patterns
• analysing facial expression, word choice or voice in online interviews
• analysing applicant or employee data bought from third parties

Word processing, spreadsheets, hosting, security tools and similar technology are excluded, provided they do not make a decision about an employment benefit.

An agent is anyone acting on the employer’s behalf, directly or indirectly, to carry out a function traditionally exercised by the employer, which can include recruitment, screening, hiring, promotion and decisions on pay, benefits or leave, including where done through an automated-decision system. An agent is itself an employer under FEHA.

What it requires

No discriminatory systems or proxies. It is unlawful for an employer to use an automated-decision system or selection criteria, including a qualification standard, test or proxy, that discriminates against an applicant or employee on a basis FEHA protects, subject to the usual defences. A proxy is a characteristic or category closely correlated with a protected basis.

Anti-bias testing as evidence. Relevant to any such claim or defence is evidence, or the lack of evidence, of anti-bias testing or similar proactive efforts to avoid discrimination. The quality, efficacy, recency and scope of the effort, its results and the employer’s response all count. This appears throughout the regulations, including in the sections on specific protected characteristics.

Medical and psychological inquiries. An assessment delivered through an automated-decision system, such as a test, question or puzzle game, can be an unlawful medical or psychological inquiry before a job offer if it elicits information about a disability.

Criminal history. The bar on considering criminal history before a conditional job offer expressly covers doing so through an automated-decision system, as well as through applications, background checks and internet searches.

Records. Personnel and employment records, including selection criteria and automated-decision system data, must be kept for four years, up from two. Automated-decision system data covers data used in or resulting from the system and data used to develop or customise it for the employer.

Key dates

• 27 June 2025. Approved and filed with the Secretary of State.
• 1 October 2025. The regulations take effect.

How it relates to other frameworks

California now has two regulators addressing the same tools from different directions. The CPPA’s CCPA ADMT regulations set privacy duties of notice, opt-out, access and risk assessment from 2027. These regulations set the discrimination standard and the record. An employer using AI in hiring in California needs to meet both.

Elsewhere, NYC Local Law 144 mandates a bias audit and publication, which California does not, while Illinois makes discriminatory effect and missing notice violations. Taken together, the US approach to AI in employment is converging on regular testing for disparate impact, notice to candidates and a durable record, whether or not a particular law mandates each piece.

Working with it in Aicura

US state framework content is available in Aicura on the Enterprise tier and is switched on in Settings under Frameworks. The AI Register records each system used in recruitment, screening and other employment decisions, and Vendors tracks the agents and suppliers acting for you, with due diligence and the contract in force. Impact assessments set out who is affected, the rights involved and the mitigations, and the Evidence Vault, where Aicura seals timestamped snapshots so anyone can confirm they have not been changed, keeps one of each assessment and change, which helps when four years of records are expected. Aicura gives guidance for your own judgement and your advisers’. It does not run anti-bias testing on your hiring tools or decide whether a tool discriminates.