On September 30, 2026, California Governor Gavin Newsom signed a package of AI and worker-protection bills, including SB 947, which requires human review of certain automated employment decisions. The bills focus on discipline, termination, displacement, and surveillance, not hiring specifically, but they point to where regulators are heading: AI can assist employment decisions, and a person must stay accountable for them. Employers and vendors should build that into recruiting workflows now.

Note: This is general information, not legal advice.

What Did California Sign?

According to the governor’s office announcement, the package includes:

  • SB 947 (Sen. McNerney), “Employment: automated decision systems,” which bars employers from relying solely on AI for disciplinary or termination decisions so that real people review automated employment actions. The governor’s release describes this requirement; the California Labor Federation’s statement on the same page names SB 947.
  • SB 951 (Sen. Reyes), “Employment: technological displacement: notice,” which requires employers to disclose when an AI system causes a mass layoff, relocation, or termination.
  • AB 1331 and AB 1883, which address workplace surveillance.

A law firm summary (Gibson Dunn, October 2026) reports a July 1, 2027 effective date for SB 947. Confirm against the enacted bill text before relying on it.

Do These Laws Apply to Hiring?

Not directly, based on the governor’s description. The listed bills address discipline, termination, displacement, and surveillance. Treat them as a signal, not a hiring rulebook. Existing rules already reach hiring: anti-discrimination law applies to automated tools, and the adverse impact and four-fifths rule framework remains the common test for screening outcomes. State regimes such as Colorado’s AI act add further obligations.

What Does Human Review Look Like in Recruiting?

A defensible human-in-the-loop design usually has four parts:

  1. AI recommends, a person decides. Scoring or ranking informs a recruiter, who owns the decision to advance or reject. See what human-in-the-loop recruiting means.
  2. Reviewers have the context and authority to disagree. A rubber-stamp approval is not meaningful review.
  3. Decisions are logged. Record what the system recommended, what the human decided, and why.
  4. Outcomes are monitored. Track selection rates across groups and investigate disparities.

These practices fit how agentic systems should operate; the AI agents vs copilots guide explains how autonomy levels differ, and agentic screening covers where screening decisions sit.

What Should Employers and Vendors Do Now?

  • Inventory your automated tools. List every system that scores, ranks, screens, or recommends employment actions, in hiring and beyond.
  • Map each tool’s decision point. Identify where a human reviews and where the tool acts alone.
  • Ask vendors for documentation. Request explanations of how recommendations are produced and what audit logs exist.
  • Watch the timeline. Review effective dates and talk to counsel.

Joveo’s approach to agentic recruiting keeps recruiters in control of decisions while agents handle the repetitive work of advertising, screening support, and follow-up.

Frequently Asked Questions

What is California SB 947? 

A bill signed September 30, 2026 that requires human review of automated employment decisions such as discipline and termination, per the governor’s announcement.

Does SB 947 apply to hiring decisions? 

The governor’s announcement describes disciplinary and termination decisions, not hiring. Consult counsel on how your tools are covered.

What is SB 951? 

A bill requiring employers to disclose when an AI system causes a mass layoff, relocation, or termination.

When do the rules take effect? 

A secondary source reports July 1, 2027 for SB 947; verify against the enacted text.

Does using AI in recruiting violate these rules? 

These bills do not prohibit AI. They emphasize human accountability.