California's Assembly Appropriations Committee holds a suspense hearing today — the make-or-break gate before any bill can reach a floor vote. On the docket: SB 947, the "No Robo Bosses Act of 2026," written by Sen. Jerry McNerney (D, SD-5). The bill would bar employers from relying solely on an automated decision system to fire or discipline a worker, and would require a plain-language written notice to any employee affected by an AI-assisted decision. It passed the California Senate 29-9 in May. Its predecessor, SB 7, cleared both chambers in 2025 before Gov. Gavin Newsom vetoed it in October 2025, calling it overly broad. No federal AI law exists, so California's bills carry weight far beyond the state.
1. Workers Need a Human Override (Sen. Jerry McNerney, California Federation of Labor Unions)
AI is already making high-stakes calls about workers' jobs, with no safeguard and no appeal.
No rules exist yet for when an algorithm fires you. Sen. McNerney says that's the gap SB 947 fills: "Employers are increasingly using AI to boost productivity and achieve cost-savings, but there are no safeguards in place to prevent harm to workers." The bill doesn't ban AI in the workplace. It says you can't fire or discipline someone on an algorithm's recommendation alone — a human has to weigh in first.
AI job displacement is happening now, not someday. Lorena Gonzalez, president of the California Federation of Labor Unions, has called it "a political choice, not an inevitability," and has demanded "human review and oversight of any decision made by a machine that impacts a worker's job and paycheck." The labor package before the committee also includes SB 951, which would require 30 days' notice before AI-driven mass layoffs.
SB 947 takes a narrower shot at the governor's complaint. The 2025 bill required pre-use notice for a wide range of employment decisions. McNerney's version cuts it to discipline and termination only, with post-use notice rather than pre-use — directly targeting what Newsom called "unfocused notification requirements."
2. The Bill Is Still Too Broad (California Chamber of Commerce, Chamber of Progress)
Defining "automated decision system" too loosely would freeze legitimate uses of AI — and create a compliance nightmare.
SB 947's definition of "automated decision system" is the problem. The California Chamber of Commerce and the Chamber of Progress argue the language is overbroad — that restrictions on predictive AI would sweep in tools financial institutions use to flag fraud risk, a use that has nothing to do with firing workers. They've called it a "chilling effect" on AI development in the state.
This is an unfunded mandate for local governments. Public employers have flagged the bill as a compliance cost they'd have to absorb with no state money to cover it — meaning cities and counties would need to overhaul HR processes at their own expense.
The business argument is that California is legislating alone. With no federal AI law in place, state-by-state rules create a patchwork. Critics say employers will simply adopt AI-based HR tools through vendors in other states, and California workers will end up with less transparency, not more.
3. Study First, Then Act (Gov. Gavin Newsom)
Newsom supports worker protections in principle — but he's already vetoed one version and hasn't committed to this one.
SB 7 died on Newsom's desk for being too imprecise. His October 2025 veto message said the bill imposed "overly broad restrictions" and notification requirements that were "unfocused" — applying to even "innocuous" AI tools. He said it failed to target the specific scenarios where employers actually misuse AI against workers.
His alternative is a study, not a law. On May 21, 2026, Newsom signed Executive Order N-6-26, directing state agencies to research AI's labor market impact and recommend WARN Act updates. Gonzalez's response was direct: "It's not enough to just study the issue — we have to take action now."
The redesign targets Newsom's specific objections. Whether that's enough to win his signature, he hasn't said. He hasn't committed publicly to signing or vetoing SB 947.
Where This Lands
Today's vote is the first gate, not the last. Labor unions and Sen. McNerney say the bill fills a real gap: no existing law stops a company from using an algorithm alone to fire someone. The California Chamber of Commerce and tech groups say SB 947's definitions are still too loose — that a California-only patchwork creates compliance chaos without fixing anything. Newsom vetoed the predecessor and signed a study in its place; he hasn't said where he stands on SB 947. If the bill clears the committee today, it heads toward a floor vote and then to the governor's desk.
Sources
- https://sd05.senate.ca.gov/news/ca-senate-approves-no-robo-bosses-act-2026-ensure-human-oversight-ai-workplace
- https://legiscan.com/CA/bill/SB947/2025
- https://www.crowell.com/en/insights/client-alerts/california-sb-947-no-robo-bosses-act-new-proposed-guardrails-on-use-of-automated-decision-systems-in-employer-discipline-and-termination-decisions
- https://calaborfed.org/press-releases/california-labor-unions-demand-transparency-and-human-oversight-of-artificial-intelligence-with-new-legislation/
- https://calaborfed.org/press-releases/california-labor-president-lorena-gonzalez-responds-to-governors-ai-executive-order/
- https://www.fisherphillips.com/en/insights/insights/california-governor-vetoes-no-robo-bosses-act
- https://pro.stateaffairs.com/ca/ai-tech/lawmaker-introduces-no-robo-bosses-act
- https://progresschamber.org/resources/letter-to-ca-lawmakers-oppose-overbroad-automated-decision-system-restrictions-in-the-workplace-sb-947/
- https://www.transparencycoalition.ai/news/ai-legislative-update-august7-2026
- https://www.cdflaborlaw.com/blog/governor-newsom-signs-executive-order-to-confront-economic-impacts-of-ai/
- https://www.lw.com/en/insights/california-assumes-role-as-lead-us-regulator-of-ai
- https://perkinscoie.com/insights/update/navigating-growing-landscape-state-ai-employment-bills-and-laws-what-employers-need