TL;DR
California's "No Robo Bosses Act" makes the state the first in the US to ban AI-only firing decisions, forcing human review before any termination driven primarily by automated systems.
What happened
- Governor Newsom signed SB 947, the "No Robo Bosses Act," on October 2, 2026, one day after Trump hosted AI executives at the White House.
- The law prohibits employers from relying solely on AI to fire or discipline workers; if AI is the primary driver, a human must corroborate the decision with manager evaluations, work samples, or personnel records.
- Workers get written notice when AI was primarily used, including what data the system processed and a named human contact who can explain the decision.
- Takes effect July 2027, giving employers roughly nine months to redesign termination workflows.
- Newsom vetoed an earlier version last year; the reintroduced bill dropped advance-notice requirements and removed gig-worker protections that had drawn opposition from Uber and Lyft.
Why it matters
- One in four American managers already uses AI to decide who to keep or cut, per an HR Dive survey, making this a live operational issue, not a hypothetical.
- The law extends beyond firing: companion bills restrict AI emotional-state monitoring, ban workplace surveillance in bathrooms, require layoff notices to disclose technological displacement, and bar AI from replacing clinical judgment in health care.
- California's scale forces national compliance: any employer with California workers must build human-review checkpoints into automated HR systems, raising costs and complexity for platforms sold nationwide.
- The "primarily" standard is contested: the Chamber of Progress, backed by major tech companies, calls the threshold too vague, signaling litigation risk before the law even takes effect.
- The package directly counters federal deregulation: Newsom explicitly rejected Trump's voluntary "morally binding" AI safety agreement and ordered state agencies to keep using "artificial intelligence" rather than Trump's preferred term "Super Intelligence."
What to watch next
- Legal challenges from employers or tech lobbies testing what "primarily" means in practice, which could blunt or sharpen the law before July 2027.
- Other states copying the California template, as has happened repeatedly with California privacy and labor law, which would push this standard toward a de facto national baseline.
- Federal preemption attempts: if Congress or the White House moves to block state AI employment rules, California's package becomes a flashpoint in the broader federal-versus-state AI governance fight.
Originally published on Present of AI, a daily source-linked AI news timeline. Read the full timeline or browse the open dataset.