California SB 947 Is Signed: What the No Robo Bosses Act Means for Staffing and Consulting Firms

California SB 947 Is Signed: What the No Robo Bosses Act Means for Staffing and Consulting Firms

Governor Newsom signed California SB 947, the No Robo Bosses Act, on September 30, 2026. The bill passed the Senate 28–10 and the Assembly 53–14 — not a squeaker, a mandate. It is now law, effective July 1, 2027, with a $500-per-violation penalty, punitive damages, and attorney fees.

If you run a staffing agency, a consulting firm that designs HR workflows, or any professional services operation that helps clients manage people decisions with AI — this law is directly about your work.

SB 947 is now law. For a step-by-step compliance checklist, see California SB 947 Signed: What Staffing and Consulting Firms Must Do Now.

What SB 947 Actually Requires

The No Robo Bosses Act is not a ban on AI in HR. It is a human-review mandate.

Under SB 947, any employer using AI in a way that materially influences a termination, suspension, demotion, formal discipline, or performance improvement plan placement decision must have a documented human review step before that decision is executed. "AI-informed" is broader than "AI-automated" — you do not need to be running a fully automated firing process to be in scope. If AI output was a material input to the decision, the human review requirement applies.

Covered decisions: termination, suspension, demotion, formal discipline, PIP placement based on AI-generated metrics.

Not covered: scheduling, compensation modeling, general workload reporting. The trigger is a consequential employment action, not any AI use in HR.

The effective date is July 1, 2027. The implementation window is nine months. That sounds like time. It is not as much time as it sounds — especially if your firm has designed these workflows for multiple clients.

Why Staffing Agencies Are the Highest-Risk Category

Staffing agencies are not peripheral to SB 947 — they are squarely in scope.

Staffing agencies routinely use AI for resume screening, candidate ranking, interview scoring, and placement speed metrics. Many of those outputs feed directly into client decisions about placement, continued engagement, and contract termination. If an AI-ranked candidate list informs a client's decision to end a worker's assignment, and no documented human review occurred before that decision, that is a potential SB 947 violation in the client's organization — and the staffing agency may have designed the workflow that made it possible.

California's general liability framework does not require the staffing agency to be the employer for liability to flow back up the chain. If your firm designed an AI-assisted HR workflow that a client ran without a human review step, the exposure trail does not stop at the client.

What to do now: Audit every AI-assisted HR workflow your firm runs for California-based clients. Map which outputs feed into employment decision chains. Document where the human review checkpoint is — or note that one does not currently exist. Build that checkpoint in before July 1, 2027.

What Consulting Firms Need to Know

If your consulting firm has advised a client on AI-assisted HR workflow design, and that workflow is now running in California without a human review requirement, your client is potentially in scope — and your engagement documentation is part of the paper trail.

Proactive step: add human review checkpoint language to every AI HR implementation deliverable you produce. Not a footnote. A clearly documented step in the workflow that specifies who reviews, using what information, before what decision is made. When a client's legal team reviews SB 947 compliance, your deliverable should show the checkpoint was designed in, not added as an afterthought.

The consulting liability angle is indirect but real. Firms that help design AI HR workflows bear responsibility for what those workflows enable downstream. The $500-per-violation structure is per violation — for an organization making hundreds of AI-assisted employment decisions monthly, the numbers scale quickly.

What Newsom's Signature Means for Your Firm

SB 947 takes effect July 1, 2027. Every California employer using AI in consequential employment decisions — and every professional services firm that designed those systems — has nine months to comply. The requirements: identify every AI-informed termination or discipline decision process, insert a documented human review step, train the supervisors responsible for that review, and update vendor agreements and client SLAs to reflect the obligation.

Staffing agencies need to update service agreements with California clients to specify human oversight requirements and clarify where SB 947 compliance obligations fall. Consulting firms need to update standard deliverable templates before engaging on any new AI HR project.

Connecticut SB 5 — the most comprehensive state AI employment law enacted to date — already requires human review for consequential AI employment decisions affecting Connecticut employees. California's signing adds weight to what is now a clear multistate standard.

Three Actions to Take Before July 1, 2027

1. Map every AI-assisted HR workflow touching California clients.

You cannot audit what you have not documented. Build a list: which workflows use AI output, which decisions those outputs feed into, and which clients or engagements are California-based. This document is the foundation of SB 947 compliance — and it is what your legal counsel will ask for if a client dispute arises.

2. Add a documented "human review" checkpoint before any termination or discipline decision.

One step in the workflow. A named person or role. Documented evidence that the review occurred before the decision was executed. This is the core requirement of SB 947, and it is the core of reasonable AI HR governance regardless of what any specific law requires.

3. Update client engagement letters and vendor agreements.

If you design AI HR workflows for clients, your standard engagement language should now specify that AI-informed employment decisions require documented human review before execution. For existing California clients on current engagements, send an addendum now. For new clients, add it to the standard template before the next engagement starts.

The risk of inaction is asymmetric. Adding a human review checkpoint to your AI HR workflow costs process design time. Not adding one, and then having your client face a $500-per-violation claim under SB 947, costs considerably more — and the liability trail leads back to whoever designed the workflow.

One Action This Week

Draft the human review checkpoint language for your standard engagement template. One paragraph. It names the decision types covered (termination, suspension, formal discipline, PIP placement), specifies that AI output used as a material input requires a named human reviewer, and states that the review is documented before the decision is executed.

Draft it now so your clients are not waiting when their legal teams start asking compliance questions. The compliance clock — July 1, 2027 — is running.

---

Sources: California Legislative Information, SB 947 bill text and vote record (leginfo.legislature.ca.gov); Brief #195 (Sep 22 sweep), Research Analyst, The Crossing Co; California Privacy Rights Act, ADMT framework. For the multi-state pattern, see Colorado ADMT Compliance 2027 and Connecticut SB 5.