SB 947 was signed on September 30, 2026. Beginning July 1, 2027, California employers may not rely solely on an automated decision system to discipline or terminate a worker. When an employer primarily relies on an automated output, the law requires human corroboration, a written post-use notice, and access to a meaningful description of the employee data used.
Source checked October 1, 2026 · CHAPTER 859, STATUTES OF 2026 · AI + EMPLOYMENT DECISIONS + JOB QUALITY · Official source ↗ · AI-assisted
WDB decision strip
| Status | Action | Primary owners | Key dates | Impact |
|---|---|---|---|---|
| SIGNED LAW | REVIEW AI-DRIVEN EMPLOYMENT WORKFLOWS BEFORE JULY 1, 2027 | Employer Services + Host-Agency HR/Legal + AI/Job Quality Leads | Operative 7/1/27 | Employer services · Public employers · Job quality · AI governance |
The bottom line
SB 947 is not a new WDB program mandate. It is an employer rule with two practical workforce-system implications. First, public agencies are included in the law’s employer definition, so a WDB’s host agency may need its own HR and legal review if automated tools are used in discipline or termination. Second, employer-facing WDB staff now have a concrete reason to treat responsible AI use as part of job quality and business modernization, while referring compliance questions to qualified counsel.
What changed
| Change | What the law does | WDB implication |
|---|---|---|
| No AI-only discipline or termination | An employer may not rely solely on an automated decision system when making a disciplinary or termination decision. | Employer-facing staff should avoid presenting AI adoption as a simple efficiency play. Human decision responsibility remains central. |
| Human corroboration | If an employer primarily relies on automated output, a human must corroborate the decision using relevant information. If the output cannot be corroborated, or is found inaccurate, incomplete, or misleading, it cannot be used for the decision. | AI procurement and implementation conversations should include the human-review workflow, not only the tool. |
| Post-use notice | An affected employee must receive a stand-alone written notice when the employer primarily relied on an automated system, including confirmation of human review and a human contact for more information. | Employers may need new documentation and communication procedures around automated employment tools. |
| Employee data access | Employees may request a meaningful, objective description of their own data primarily used by the automated system in the decision. | Data governance becomes part of responsible workforce technology use. |
| Protected rights | The law also bars specified uses of automated systems to infer protected status or predict and take adverse action because a worker exercises legal rights. | WDBs can frame responsible AI as both a technology and job-quality issue without attempting to give legal advice. |
Operational considerations for WDBs
- Separate technical assistance from legal advice. Employer-services teams can explain that California has new guardrails for automated employment decisions, then refer compliance questions to counsel or appropriate state resources.
- Add human-review questions to AI-readiness work. When discussing AI with employers, ask who verifies automated outputs, what source data can be checked, and who owns the final decision.
- Review the WDB host agency’s own use of employment technology. If county, city, nonprofit, or other host-agency HR systems use automated outputs for discipline or termination, route the issue to HR, procurement, privacy, and legal teams.
- Connect the policy to job quality. Employer engagement can include transparency, worker voice, privacy, and clear appeal or review processes as part of responsible technology adoption.
- Train staff on the boundary. WDB staff should know the basic rule and the July 1, 2027 operative date, but should not determine whether a particular employer’s system is legally compliant.
Required / local choice / good practice / watch out
| Label | How to apply it |
|---|---|
| Required | Covered employers must comply with the statute’s automated-decision requirements beginning July 1, 2027. The law assigns enforcement to the Labor Commissioner and public prosecutors, not to local WDBs. |
| Local choice | A WDB may decide whether to incorporate responsible AI and job-quality questions into employer services, sector work, procurement guidance, or staff training. |
| Good practice | Treat the human review process, source data, notice, and accountability structure as part of an AI implementation plan rather than an afterthought. |
| Watch out | Do not describe SB 947 as a ban on workplace AI. It regulates specified uses of automated decision systems and does not make WDB staff employment-law compliance advisers. |
Source + Trust Record
| Primary authority | California Legislative Information – SB 947 |
|---|---|
| Signing confirmation | Governor of California – September 30 AI and worker-protection package |
| Current status | CHAPTER 859, STATUTES OF 2026 |
| Source checked | October 1, 2026 |
| What we verified | SB 947 is Chapter 859, Statutes of 2026. The enrolled law is operative July 1, 2027 and prohibits sole reliance on an automated decision system for discipline or termination while adding human corroboration, notice, data-description, anti-retaliation, and enforcement provisions. |
| Important limitation | The statute regulates employers. Workforce Wonkery’s WDB suggestions are operational analysis and are not legal, HR, privacy, procurement, or compliance advice. |
Official sources control. Workforce Wonkery is AI-assisted and does not receive human legal or compliance review.
Leave a Reply