Last year, we wrote about the growing trend of managers using artificial intelligence to make employment decisions and the significant legal risks that can arise when employers rely too heavily on AI-generated recommendations. At the time, our advice was straightforward: AI can be a useful tool, but it should never replace human judgment, particularly with regard to tangible employment actions. Now California has taken that principle one step further by turning it into law.

On September 30, 2026, California Gov. Gavin Newsom signed the “No Robo Bosses” Act (SB 947), which takes effect July 1, 2027, making California the first state in the nation to require meaningful human oversight of certain AI-assisted employment decisions. The law reflects a growing concern among lawmakers and regulators that employers are increasingly relying on AI tools to make high-stakes employment decisions without sufficient human review.

A Quick Refresher: The Risks Haven’t Changed

As discussed in our previous article, some employers have incorporated the use of AI tools for hiring, promotions, discipline, layoffs, and terminations. However, when human oversight is not used as a backstop, AI systems can replicate and amplify historical biases embedded in their training data. Notable examples include Amazon’s abandoned recruiting tool, which reportedly favored male applicants, and the ongoing litigation against Workday, Inc. involving allegations that AI-driven applicant screening disproportionately harmed applicants with certain protected characteristics. Federal agencies, including the Equal Employment Opportunity Commission (EEOC), also have warned employers repeatedly that they may be liable for discriminatory outcomes produced by AI tools, even if those tools were developed by third-party vendors.

The main takeaway for employers is that using AI does not eliminate responsibility for employment decisions: Employers cannot hide behind the automated nature of these tools.

California’s Answer: Keep Humans in the Loop

The No Robo Bosses Act takes direct aim at automated employment decision-making. Under the new California law, employers are prohibited from relying solely on an automated decision system to discipline or terminate employees. Instead, a human reviewer must evaluate and corroborate the decision using relevant information, such as performance evaluations, personnel records, peer reviews, and other workplace data.

The law also gives employees new transparency rights. Employees must be notified when an automated decision system played a role in a disciplinary or termination decision and may request a meaningful description of the employee’s own data used by the system. Additionally, employers cannot use an automated system if a human reviewer cannot independently verify or corroborate the system’s “output,” which can be any information, data, assumptions, predictions, scoring, recommendations, decisions, or conclusions that the system generated. Employers may not retaliate against an employee for requesting a copy of the output relied on in the employment decision.

Why Employers Everywhere Should Pay Attention

While the No Robo Bosses Act applies only to California employers, employers outside the state should not dismiss it as an isolated development. California is often an early adopter in workplace regulation, particularly in emerging areas involving technology and employee rights. Moreover, California’s action comes amid a broader trend of state-level AI regulation. As we blogged about earlier this year, Connecticut’s new law requires that employers explain when technology influences employment decisions. Colorado has continued refining its AI governance framework, New Jersey lawmakers are considering workplace AI restrictions, and New York legislators have explored audit and disclosure requirements for AI systems used in employment decisions.

Further, California’s new law underscores several best practices employers should already be implementing:

• Ensure that AI tools are used to inform decisions rather than make them.
• Maintain meaningful human review of any AI-generated recommendation involving hiring, promotion, discipline, termination, or layoffs.
• Document the human review, including the underlying data and non-AI factors considered in each employment decision.
• Train managers on the capabilities and limitations of AI tools.
• Conduct periodic audits of AI systems to identify potential bias or disparate impact.
• Review existing AI governance policies and update them as laws continue to evolve.

And, for California employers, the compliance clock is already running. Before July 1, 2027, California employers should:

1. Inventory every tool that scores, ranks, or flags employees for discipline or termination.
2. Designate and train human reviewers to corroborate system outputs against personnel records and other workplace data.
3. Prepare the § 1524 written-notice template for employees affected by AI-assisted decisions.
4. Set up a process, and handbook language, for employee requests to describe their data used by the system.
5. Confirm reviewers can independently verify each tool’s output, and that requests for it are never met with retaliation.

Looking Ahead

In our prior article, we asked: “What could go wrong?” California’s response appears to be “quite a bit,” unless humans remain accountable for workplace decisions. As AI becomes increasingly integrated into day-to-day management functions, employers should expect additional regulation, heightened scrutiny, and continued litigation testing the boundaries of AI-assisted employment practices. The safest approach remains the same today as it was when we first wrote on the topic: Use AI as a tool, not a substitute for sound managerial judgment. If you have questions about how the No Robo Bosses Act or other state AI laws affect your workforce decisions, please contact the authors.