Published: October 1, 2026 · Last updated: October 1, 2026
TL;DR: Governor Gavin Newsom signed the No Robo Bosses Act on September 30, requiring a human to sign off before an employer can fire or discipline a worker based on an automated system, after vetoing an earlier version of the same bill.
California employers can no longer let an algorithm make the final call on firing someone. Governor Gavin Newsom signed SB 947, the No Robo Bosses Act, on September 30, reversing course after he vetoed an earlier version of the same idea, SB 7, the year before.
State Senator Jerry McNerney reintroduced the bill in February 2026 after that veto, and this time it made it to Newsom’s desk and got a signature. The law takes effect July 1, 2027, and it requires employers using automated decision systems for termination or discipline to get independent human corroboration before acting on whatever the system recommends. Workers who are let go or disciplined based on an automated system also have to receive timely, plain-language written notice, including contact information for a human reviewer and details on how to access the data the system used about them.
The bill goes further than just requiring a rubber stamp. It explicitly bars employers from using automated systems to predict employee behavior or infer protected characteristics, like disability or pregnancy status, from unrelated data points. That’s aimed at a specific failure mode of workplace AI tools: a system that’s never told to consider someone’s medical leave history might still end up penalizing them for it indirectly, through attendance patterns or performance dips the algorithm wasn’t built to contextualize.
Penalties for violations run $500 per incident, with the door open for punitive damages and attorney’s fees on top of that. For a large employer running automated systems across thousands of workers, that adds up fast if the system is making bad calls at scale rather than the kind of one-off mistake a human manager might catch and quietly correct.
McNerney’s framing throughout the process was about restoring a check that already should have existed. Human oversight and independent verification, he argued, reduce the risk of sole reliance on systems that weren’t built to carry that kind of consequence alone. It’s hard to disagree with the principle. The harder question is enforcement: a written notice requirement only protects workers who know to read it and know what to do with a human reviewer’s contact information once they have it.
California has a pattern of setting the national template on tech regulation that other states eventually copy, from data privacy to gig worker classification. The No Robo Bosses Act is likely to follow that same arc, especially as more employers quietly fold AI-driven HR tools into hiring, scheduling, and performance review pipelines without workers ever being told a model is involved.
This is general reporting on new legislation, not legal advice. If this law affects your job, consult a licensed employment attorney.
Related: The Pentagon Just Put Elon Musk and Palmer Luckey in Charge of Imagining the Next War and Reddit Is Killing RSS Feeds This November, Blaming AI Bots for Ruining It.
Bottom Line: This law doesn’t ban AI from workplace decisions, it just puts a human back in the loop before the worst outcomes. That’s a lower bar than it sounds, and the real test comes in 2027 when employers have to actually prove they cleared it.
Follow Teck Hustlers: Facebook · X · Instagram · LinkedIn · TikTok · Pinterest
