THU, OCTOBER 01, 2026
Independent · In‑Depth · Practitioner‑Tested
✎ General

California Legislated Into the Path of Federal Preemption

SB 947 bars automated systems from being the primary driver of a firing and forces employers to show the worker what data was used - signed one day after the White House started a 60-day clock on a federal AI definition.

By AIToolsRecap October 1, 2026 6 min read 9 views
Home › Articles › General › California Just Banned AI-Only Firings

What the Law Does

California Governor Gavin Newsom signed SB 947, the No Robo Bosses Act, on 30 September 2026.

It prohibits employers from using automated decision-making systems as the primary driver of a decision to fire or discipline a worker.

It requires:

  • Human review that corroborates automated findings against personnel files, manager evaluations and peer reviews
  • Written notice to the affected worker detailing the AI data used in the decision
  • A named human contact who can explain the decision

Note the wording: primary driver, not "any use". An employer can still run automated systems over workforce data. It cannot let the output be the thing that ends someone's job without a human having independently checked it.

What Is Not Yet Clear

Reporting has not confirmed the effective date, the penalties, or exactly which employers are covered. Those three things determine whether this has teeth, and we are not going to guess at them. If you have a compliance interest, read the enrolled bill text rather than the coverage, including ours.

The Compromise That Got It Signed

Newsom vetoed an earlier version in October 2025, citing overly broad pre-notification rules.

The version he signed removed the advance notice requirements — and eliminated protections for gig workers.

That second change deserves more attention than it is getting. Algorithmic management is not evenly distributed across the workforce. The people most likely to be hired, scheduled, rated, disciplined and deactivated by software are precisely platform workers. A law restricting AI-driven termination that excludes gig workers protects the employees least exposed to the practice and leaves out the ones most exposed to it.

That is what made it passable. It is also the main limit on what it achieves.

The Timing Is Not a Coincidence Worth Ignoring

On 29 September, the White House directed federal agencies to refer to AI as "super intelligence" and gave the president's science adviser 60 days to draft legislative language defining the term, including whether it should modify, expand or replace existing legal definitions of AI. That clause is widely read as the groundwork for federal preemption of state AI law.

On 30 September, California signed a state AI law anyway.

Whatever else SB 947 is, it is also a jurisdiction legislating into the path of a preemption effort, from the state where most of the companies in question are headquartered. We covered the executive order and the accord here.

The next 60 days decide whether laws like this one survive as written.

What It Means If You Build HR or Workforce Tools

If your product scores, ranks, flags or recommends action on employees, and any customer operates in California, this reaches you — not because the law regulates your software, but because your customers now need things from it:

  • An audit trail of what the model saw. The written-notice requirement means the employer has to tell the worker what data drove the decision. If your product cannot produce that, your customer cannot comply.
  • A record of human review. Corroboration has to be demonstrable, not asserted.
  • Output framed as input, not verdict. A product whose interface presents a conclusion invites exactly the use the law prohibits. One that surfaces evidence a manager weighs does not.

The third point is a design question rather than a legal one, and it is the one worth acting on first.

FAQ

What is the No Robo Bosses Act?

California SB 947, signed 30 September 2026, barring employers from using automated decision-making systems as the primary driver of firing or disciplinary decisions.

Does it ban AI in HR entirely?

No. It bars automated systems from being the primary driver of termination or discipline, and requires corroborating human review plus written notice to the worker.

Does it cover gig workers?

No. Protections for gig workers were removed from the version that passed.

When does it take effect?

Not confirmed in reporting. Check the enrolled bill text.

Why did Newsom sign it this time?

He vetoed an earlier version in October 2025 over overly broad pre-notification rules. The revised bill dropped advance notice requirements and gig worker protections.

Tags
AI News2026Productivity
⚑

Spot an inaccuracy?

We verify facts before publishing and correct errors promptly. If something in this article is wrong or outdated, let us know.

Report an error →
💡 AI Tools prompts
Prompt Guide
Best Claude AI Prompts for SEO (2026) — Content, Technical, and Comparison SEO
Claude Sonnet 5 and Opus 5 are strong for SEO work that requires writing quality, structured analysis, and long-form content generation. With 1M context, Claude can analyse an entire site's content structure, compare competing pages, and write complete article drafts in one session. These prompts cover the full SEO workflow: keyword research synthesis, content briefs, on-page optimisation, meta descriptions, technical audit interpretation, and comparison content that ranks above AI Overviews.
Get Prompts →
Prompt Guide
Best ChatGPT Prompts for SEO (2026) — GPT-5.6 and Browse
ChatGPT with GPT-5.6 Sol and Browse enabled is a capable SEO research tool — it can search the live web, analyse SERP results, and synthesise content briefs in a single session. GPT-5.6 Terra at $2.50/M offers a cost-efficient option for high-volume SEO content generation. These prompts are optimised for ChatGPT Plus with Browse, the ChatGPT Work product for larger projects, and the OpenAI API with web_search tool enabled.
Get Prompts →
Prompt Guide
Best Claude Opus 5 and Sonnet 5 Prompts for Writing (2026)
Claude Opus 5 and Sonnet 5 consistently produce the highest-quality long-form writing of any AI model in July 2026 — a lead documented across writing benchmarks and user testing since Claude 3 Opus. With 1M context and 128K output on Opus 5, Claude can write book chapters, complete reports, and long-form content without truncating. Sonnet 5 at $2/$10/M (intro through August 31) is the best value writing model available. These prompts are optimised for claude.ai Pro/Max, Claude Cowork, and the API.
Get Prompts →