Blog: The California Workplace Advisor

California employers using artificial intelligence and automated technology in the workplace have new compliance obligations to prepare for.

On September 30, 2026, Governor Gavin Newsom signed four measures addressing AI and technology in the workplace: SB 947, SB 951, AB 1331, and AB 1883. Together, the laws address automated employment decisions, AI-related workforce reductions, and workplace surveillance.

Three of the new laws—SB 951, AB 1331, and AB 1883—take effect January 1, 2027. SB 947’s requirements governing automated decision systems become operative July 1, 2027. Employers should begin reviewing their technology and practices now so they are prepared when the new requirements become effective.

The takeaway for employers—particularly startups and other companies adopting new technology quickly—is straightforward: understand which tools you are using, what those tools actually do, and whether their outputs influence employment decisions or monitor employees.

At a high level:

• SB 947 says employers generally cannot let an automated system make a disciplinary or termination decision by itself. A person must review the relevant facts and make an independent decision.
• SB 951 adds new disclosure requirements when AI or other automated technology substantially contributes to certain workforce reductions covered by California’s WARN Act.
• AB 1331 restricts employers’ use of surveillance technology to monitor employees in workplace bathrooms.
• AB 1883 prohibits employers from using certain AI surveillance tools to collect employee neural data or recognize or infer employees’ emotional states.

These laws establish new boundaries around the use of AI and automated technology in the workplace, particularly when technology is used to monitor employees or inform employment decisions.

Begin With an Audit of Your Tech Stack
In terms of the right next step, shelve the lengthy AI policy and conduct a straightforward audit of the software and platforms your company uses to hire, manage, monitor, evaluate, or terminate California-based employees.

1. Inventory the Technology
Identify every platform or tool that touches applicants or employees.

Think beyond recruiting software. Consider systems used for scheduling, performance management, productivity monitoring, workforce planning, promotion, discipline, termination, attendance, and security.

Do not assume you already know where AI is being used. Many platforms now incorporate automated or AI-powered functionality into products originally purchased for entirely different purposes. Talk to your vendors and involve them in this process.

2. Identify What Each Tool Actually Does
Determine what information the system collects and what it produces.

Does it generate a score, ranking, recommendation, flag, prediction, alert, or other automated output?

Does it monitor behavior, productivity, location, communications, biometrics, or other employee activity?

The label a vendor puts on the product matters less than its actual functionality.

3. Map the Technology to Employment Decisions
Identify which employment decisions are influenced by the tool’s output.

A system does not have to make the final decision itself to create compliance risk. Employers should understand whether decision-makers receive or rely upon automated information when making decisions involving hiring, promotion, performance, scheduling, discipline, termination, or workforce reductions.

4. Confirm Meaningful Human Review
For higher-risk employment decisions, particularly discipline and termination, identify who independently reviews the underlying facts and makes the ultimate decision.

Human review should be more than approving an automated recommendation.

The reviewer should have sufficient information to evaluate the basis for the system’s output and reach an independent conclusion.

Employers should therefore ask a deceptively simple question:

If we removed the algorithm’s recommendation, could the human decision-maker independently explain why this decision was made?

5. Document the Process
Employers should be able to explain what the technology did, what information it relied upon, what the human decision-maker reviewed, and how the ultimate decision was reached.

That documentation may become increasingly important if an employment decision is later challenged or the company needs to demonstrate compliance with California’s requirements.

The goal is to understand what the technology is doing before its recommendation becomes part of an employment decision.

Companies that begin mapping their technology now will be better positioned not only to comply with these new requirements, but also to respond as workplace AI regulation continues to develop.

For more information, contact China Daly in the San Francisco office of Hirschfeld Kraemer LLP. She can be reached at 415-835-9027 or cdaly@hkemploymentlaw.com.