California’s SB 947, the No Robo Bosses Act, bars employers from relying solely on AI to fire or discipline staff, mandating a human reviewer and written notice that AI was primarily used. It grants employees the right to know what data the system used and who can explain the decision. The bill was introduced in 2025 and is backed by labor unions seeking to curb AI-driven adverse actions in the workplace.
What Happened
California Governor Gavin Newsom signed SB 947, the “No Robo Bosses Act,” which blocks Golden State employers from relying solely on artificial intelligence to fire or discipline workers. The law prevents businesses statewide from exclusively using “automated decision‑making systems” in matters of employee discipline and termination, and restricts the use of AI as the principal tool in those decisions.
Under SB 947, an employer that relies “primarily” on AI output must have a human reviewer corroborate the decision using additional information such as managerial evaluations, peer reviews and personnel files.
Affected employees gain three concrete rights. They must receive written notice that AI was “primarily used” in their termination or disciplinary decision, a description of the employee data the system used, and a human point of contact who can further explain it.
State Senator Jerry McNerney, the bill’s author, told CNBC: “No worker should ever be fired or disciplined by a machine, AI or not.” He added: “Artificial intelligence systems have the potential to boost productivity, but they’ve also made errors and misjudgments and exhibited bias.” McNerney emphasized that “AI must remain a tool controlled by humans, not the other way around.” The Democrat first introduced the act in 2025, following a concerted push by organized labor to erect guardrails around management’s ability to use AI in taking adverse action against workers.
Lorena Gonzalez, president of the California Federation of Labor Unions, AFL‑CIO and lead sponsor of the No Robo Bosses Act, said in a statement after the signing: “When working people organize, we get results.” She added that California’s workers and unions are shaping the national conversation on AI in the workplace.
What This Means For You
If you manage people in California, treat this as an operational project, not a legal abstraction. Start by inventorying every tool that touches performance, attendance, scheduling or conduct data.
Ask vendors one blunt question: does this product ever recommend, flag or rank an employee for discipline? If the answer is yes, you need documentation a human reviewer can actually work from.
Build the corroboration step into the workflow before the decision is finalized, not after. A manager signing off on an AI score without new evidence probably will not meet the standard the law describes.
Keep the artifacts the bill names — evaluations, peer reviews, personnel files — in a form you can produce quickly. Those records are what separate real review from a rubber stamp.
Draft your written notice template now. It needs three elements: that AI was primarily used, what employee data fed the system, and who the named human contact is.
Choose that contact deliberately. It should be a person with authority to explain the reasoning, not a shared support inbox nobody monitors.
Your HR technology budget may shift too. Expect vendors to sell audit logs, human‑in‑the‑loop dashboards and automated notice generators as compliance features within the next few quarters.
For workers, the notice requirement creates leverage. If AI played a primary role in your discipline, you can request the data description and press the human contact on specifics.
If you employ people outside California, do not assume you are unaffected. Multi‑state employers rarely run two separate HR stacks, so these compliance features tend to spread across the whole organization.
Small businesses should pay attention as well. If you use a payroll or scheduling platform with algorithmic features, you may already be inside the law’s scope without realizing it.
Why It Matters
This is a state filling a federal vacuum. Senate Democrats’ AI safety bill was blocked in Washington, leaving workplace automation rules to Sacramento and other state capitals.
This suggests a broader pattern: as federal legislation stalls, states become the de facto rule‑writers for automated employment decisions. Other legislatures will likely copy SB 947’s notice‑and‑corroboration structure.
At its core, this is an AI governance story. A model’s output is only as trustworthy as the review process wrapped around it.
It also signals that claims of algorithmic error and bias now carry legal consequences in employment. Employers can no longer dismiss a bad automated recommendation as a harmless tooling quirk.
And for the AI industry, it draws a clearer line between decision support and decision authority. Tools that suggest are treated differently from tools that decide.
Key Takeaway
- California now requires a human reviewer to corroborate any termination or discipline decision that leans primarily on AI output.
- Affected workers must get written notice, a description of the data used, and a named human contact who can explain the decision.
- The No Robo Bosses Act follows a union‑led campaign that began with its 2025 introduction.
- Expect other states and HR software vendors to adopt the notice‑and‑human‑review model quickly.
Frequently Asked Questions
Can California employers still use AI at all in workforce decisions?
Yes. The law targets sole or primary reliance on automated systems, and it requires human corroboration plus disclosure when AI drives the outcome.
What exactly must a disciplined employee receive?
Written notice that AI was “primarily used,” a description of the employee data the system relied on, and a human point of contact for further explanation.
Leave a Reply