California Law Bars AI-Only Layoffs: Humans Must Decide

California has enacted a new law that prohibits layoffs decided solely by artificial intelligence: under the legislation, a human must have the final say before an employee is terminated on the basis of an algorithmic recommendation.

The measure directly targets one of the fastest-growing flashpoints in the US labor market — employers deploying automated systems to screen, rank, discipline and dismiss workers without meaningful human review. California, the largest state economy in the union and home to much of the AI industry itself, is now setting the terms for how that technology can touch employment decisions.

What does the law actually require?

The core obligation is a human-in-the-loop rule. A company may still use AI to inform or support layoff and termination decisions, but an automated system cannot be the deciding authority. A person — not a model — must make and own the final call.

For workers, this shifts accountability. When a dismissal goes wrong, there is a named human decision-maker and a reviewable process, rather than an opaque output from a black-box system.

Why does this matter for your market position?

If you work in California, or for a California-based employer, the law changes the practical risk profile of being managed by algorithm:

• AI can flag, score or recommend, but it cannot be the last word on your job.
• Terminations that skip human review now carry legal exposure for the employer.
• The law signals where regulation is heading — other states tend to follow California's lead on employment and technology rules.

For HR, people-operations and compliance professionals, the legislation creates immediate demand: someone must design, document and audit the human-review layer. That is a concrete skills signal — expertise in AI governance, employment law and workforce analytics becomes more valuable, not less, as automation spreads.

The national picture versus the sector detail

Nationally, employers continue to adopt AI across the hiring and separation pipeline, from résumé screening to performance ranking. California's law does not halt that adoption. It redraws the boundary: automation as adviser, human as adjudicator.

At the role level, the effects concentrate in sectors where algorithmic management is most entrenched — large-scale operations, tech-adjacent firms and any workforce managed through scoring dashboards. Workers in those environments gain a procedural safeguard; compliance and HR staff gain new workload and new leverage.

What comes next?

The first enforcement cases will define how strict the human-review requirement proves in practice — expect courts and regulators to test whether a rubber-stamp approval satisfies the law or whether genuine deliberation is required.