# California Sets a Statutory Workplace-AI Floor — Newsom Signs No Robo Bosses on Deadline

Times of AI Desk · 2026-09-30 · Policy

[https://timesof.ai/2026/09/california-newsom-ai-worker-transparency-package](https://timesof.ai/2026/09/california-newsom-ai-worker-transparency-package)

> On the final day to act, Governor Gavin Newsom signed SB 947 (No Robo Bosses), SB 951 AI-layoff notice, workplace-surveillance limits, SB 1000/AB 2713 transparency updates, and healthcare AI guardrails — while vetoing rival AI bills including AB 2575, AB 2656, and SB 903. The largest U.S. state just wrote a national compliance reference that Tuesday’s voluntary White House frontier accord never attempted.

Washington’s [voluntary Frontier Responsibilities accord](https://timesof.ai/2026/09/white-house-frontier-responsibilities-accord) sold “morally binding” self-policing with no legal force. Sacramento answered with statutes: the largest U.S. state just put workplace AI firing, layoff attribution, surveillance, and content provenance into binding law — a de facto national compliance floor for every employer and deployer with a California nexus.

Governor **Gavin Newsom** used the September 30 constitutional deadline to sign a package that the governor’s office frames as first-in-nation worker protections and stronger transparency. Same-day wires from **CNBC** (~7:29 p.m. EDT) and **Guardian/AP** (~8:17 p.m. EDT) put the labor bills at the center; the [gov.ca.gov package post](https://www.gov.ca.gov/2026/09/30/californias-nation-leading-ai-framework-just-got-stronger-governor-newsom-signs-more-first-in-the-nation-worker-protections-and-more/) dated September 30 supplies the signed-bill list. Newsom also contrasted state action with absent federal binding rules and Tuesday’s voluntary White House posture — the political frame, not a rehash of that accord as news.

## What signed

**Worker AI (labor package).** Per CNBC’s read of **SB 947** (No Robo Bosses Act, Sen. **Jerry McNerney**), employers may not rely **solely** on automated decision systems to fire or discipline, and face limits when AI is used as a **principal** tool. Where an employer relies **primarily** on AI for termination or discipline, a human reviewer must corroborate using additional information (managerial evaluations, peer reviews, personnel files), and the affected worker gets written notice that AI was primarily used, a description of the employee data involved, and a human point of contact. **Lorena Gonzalez**, president of the California Federation of Labor Unions, AFL-CIO, called the signing of SB 947, **SB 951**, **AB 1331**, and **AB 1883** a first-in-nation workplace guardrail set. CNBC notes Newsom had **vetoed an earlier No Robo Bosses version** last October over unfocused advance-notification duties; the 2026 bill dropped that pre-notification requirement and stripped gig-worker coverage.

**SB 951** (Sen. **Eloise Gómez Reyes**) requires written notice when a mass layoff, relocation, or termination is caused by an AI system — language mirrored in the governor’s package bullets and Guardian/AP. **AB 1331** and **AB 1883** address workplace surveillance (including bathroom bans on surveillance tools, per the governor’s summary). Guardian/AP also reports a ban on using the technology to predict a worker’s emotional state from biometric data.

**Transparency / provenance.** **SB 1000** (Sen. **Josh Becker**) and **AB 2713** (Asm. **Buffy Wicks**) strengthen the California AI Transparency Act and system provenance data. The package also covers watermark/metadata access, public-meeting comment authenticity (real people, not bots), deepfake/likeness updates (**SB 1111**), and related transparency/governance (**SB 1159**).

**Healthcare, attorneys, gene synthesis, campus AI.** Signed healthcare clinical-judgment bills include **AB 1979** and **SB 503**; attorney/AI limits **SB 574**; gene-synthesis customer screening **AB 1864**; public higher-ed AI training/procurement **AB 2392**. Full signed list is on the governor’s September 30 package page.

Newsom also signed an executive order directing state agencies to keep calling the technology **“artificial intelligence”** rather than the White House’s “super intelligence” rebrand — symbolism that Guardian/AP and the governor’s package both flag.

## What was vetoed (AI-adjacent)

The [governor’s legislative update for September 30](https://www.gov.ca.gov/2026/09/30/governor-newsom-issues-legislative-update-9-30-2026/) lists vetoes that include **AB 2575** (Asm. Liz Ortega — Health care services: artificial intelligence), **AB 2656** (Asm. Cottie Petrie-Norris — Public employees: notice: artificial intelligence performing service within scope of work), and **SB 903** (Sen. Steve Padilla — Mental health professionals: artificial intelligence), each with a linked veto message. Titles and veto status come from that update; this desk does not restate those veto letters without fetching each in full.

## The compliance trade

California employers — and any national deployer that cannot wall off the state — now face a statutory stack on automated discipline/termination, AI-attributed mass layoffs, workplace surveillance, and AI content provenance that federal voluntary self-audit never imposed. Parallel same-week federal texture includes the [FTC’s reported CID track against Anthropic, OpenAI, and METR](https://timesof.ai/2026/09/ftc-probe-anthropic-openai-metr-agents) and Italy’s criminal high-risk oversight frame already on the desk — different tools, same window: hard rules where Washington sold moral force.

Chamber of Progress and business lobbies urged a veto on SB 947’s undefined “primarily relies” standard (CNBC); EFF called the signing a strong worker step. Newsom left open a **special session** for further frontier rules (Guardian/AP) — the signed package is the floor that already exists, not a claim that California is done.

## Limits

- Signed-bill inventory is from **gov.ca.gov** (September 30 package page). SB 947 operational detail leans on **CNBC**; mass-layoff notice and biometric/emotion framing also appear in **Guardian/AP**. Prefer primary text/regs when implementing counsel digs in.
- Effective dates and implementing regulations are not fully spelled in the wires used here.
- Vetoes of AB 2575, AB 2656, and SB 903 are confirmed on the legislative update; rationales live in the linked veto messages, not paraphrased here without fetching each letter in full.
- Contrast with the White House voluntary accord is political framing from Newsom and this desk — not a claim that the accord was rewritten or withdrawn.
- Special-session possibility is Newsom leaving the door open to reporters, not a called session.

## Sources

- [Office of Governor Gavin Newsom: California’s nation-leading AI framework just got stronger (September 30, 2026)](https://www.gov.ca.gov/2026/09/30/californias-nation-leading-ai-framework-just-got-stronger-governor-newsom-signs-more-first-in-the-nation-worker-protections-and-more/)
- [Governor Newsom issues legislative update 9.30.2026 — signed and vetoed bills (September 30, 2026)](https://www.gov.ca.gov/2026/09/30/governor-newsom-issues-legislative-update-9-30-2026/)
- [CNBC: California Gov. Gavin Newsom bans AI ‘robo bosses’ in landmark state law (September 30, 2026)](https://www.cnbc.com/2026/09/30/california-gavin-newsom-ai-ban.html)
- [The Guardian / AP: Gavin Newsom signs laws to protect California workers from AI threat (September 30, 2026)](https://www.theguardian.com/us-news/2026/sep/30/gavin-newsom-california-ai-threat)
- [Times of AI: White House voluntary Frontier Responsibilities accord (September 29, 2026)](https://timesof.ai/2026/09/white-house-frontier-responsibilities-accord)
- [Times of AI: FTC probe of Anthropic, OpenAI, METR (September 30, 2026)](https://timesof.ai/2026/09/ftc-probe-anthropic-openai-metr-agents)