California’s No Robo Bosses Act (SB 947): What Workers Need to Know

On September 30, 2026, Governor Newsom signed Senate Bill 947, the “No Robo Bosses Act” (Stats. 2026, ch. 859). The law adds a new Part 5.5.5 to the Labor Code, Sections 1520 through 1526.7, titled “Automated Decision Systems in the Workplace.” Like most 2026 legislation, it takes effect January 1, 2027, but its rules on automated decision systems do not become operative until July 1, 2027 (Lab. Code § 1526.7).

The core idea is simple: software can help a manager, but it cannot be the manager when your job is on the line. SB 947 follows SB 7, a broader “No Robo Bosses” bill the Governor vetoed in 2025. The new version is narrower and focuses on discipline and termination, but it gives California workers real rights for the first time when an algorithm drives a write-up or a firing.

This guide explains the law from the worker’s side: what it covers, what it does not, what to do if a computer system played a role in your discipline or termination, and how SB 947 fits with the discrimination and retaliation protections you already have under the Fair Employment and Housing Act (FEHA). For the full list of 2027 changes, see our 2027 California employment law updates.

The Short Version

• No firing or discipline by software alone. An employer may not rely solely on an automated decision system when it disciplines or terminates a worker.
• A real person has to check the machine. If the employer primarily relies on the system’s output, a human must corroborate the decision with the underlying data or other supporting information. If the output can’t be corroborated, or looks inaccurate, incomplete, or misleading, the employer can’t use it.
• Some uses are banned outright. Employers may not use these systems to infer a worker’s protected status, to predict and punish workers for exercising legal rights, or to get around labor, safety, employment, or civil rights laws.
• You get notice and a look at your data. When a system was the primary basis for discipline or termination, you must receive a written notice when you’re told about the decision, and you can request a description of your own data the system used.
• Retaliation is prohibited. Your employer cannot punish you for using these rights or complaining about a violation.
• Hiring is not covered. The statute speaks to employees and to discipline and termination decisions. It does not expressly reach job applicants or hiring and recruiting decisions.

What Counts as an “Automated Decision System”?

SB 947 defines an automated decision system (ADS) as any computational process derived from machine learning, statistical modeling, data analytics, or artificial intelligence that issues simplified output, such as a score, classification, or recommendation, that is used to assist or replace human discretionary decision-making and materially impacts people (Lab. Code § 1520(c)).

That definition is broad. It is not limited to chatbots or cutting-edge AI. In everyday workplaces, tools that may fit include:

• Attendance or “points” systems that automatically flag a worker for a write-up.
• Productivity or “rate” trackers in warehouses, call centers, and delivery routes that rank or score workers.
• Software that turns customer ratings, sensor data, or camera analytics into a performance score.
• HR platforms that recommend who should be put on a performance plan or let go.

The law lists a few things that are not automated decision systems: spam filters, firewalls, antivirus software, identity and access management tools, calculators, and databases, datasets, or other compilations of data. Whether a particular tool qualifies will depend on what it does and how your employer used it.

The Core Rule: A Human Has to Own the Decision

Section 1522(b) draws two lines.

First, “solely” is off-limits. An employer may not rely solely on an ADS when making a disciplinary or termination decision. A manager who rubber-stamps whatever the dashboard says is the situation the law targets.

Second, “primarily” triggers a corroboration duty. If the employer primarily relies on ADS output, it must direct a human to corroborate the decision using the data that was collected or used to produce the output, or other relevant corroborating or supporting information. The statute gives examples of that kind of support: supervisory or managerial evaluations, personnel files, employee work product, peer reviews, and witness interviews (which may include relevant online customer reviews).

Section 1522(c) closes the loop. If the employer cannot corroborate the output, or the human reviewer concludes the output is inaccurate, incomplete, or misleading, the employer may not use it to discipline or fire anyone. In plain terms: a “bad score” that a person can’t back up with real facts is not supposed to cost you your job.

Uses That Are Banned No Matter What

Separate from discipline and termination, Section 1522(a) says an employer may not use an ADS to:

• Prevent compliance with, or violate, any federal, state, or local labor, occupational health and safety, employment, or civil rights law or regulation;
• Infer a worker’s protected status under Government Code section 12940, the core FEHA anti-discrimination provision (for example, race, religion, national origin, disability, medical condition, sex, gender identity, sexual orientation, age, or other protected categories); or
• Predict and take adverse action against a worker for exercising legal rights, including rights under state and federal employment and labor law.

That last item matters for workers who speak up. A system that flags people likely to complain, organize, take protected leave, or report safety problems, and then feeds that into discipline, is exactly what the law forbids.

Your Notice and Data Rights

Written notice. Under Section 1524, an employer that primarily relied on an ADS to discipline or fire a worker must give that worker a written “postuse” notice at the time it informs the worker of the decision. The notice must be in plain language, delivered as a separate, stand-alone communication (email, a link, or another written format), and in the language the employer normally uses to communicate with employees. It must tell you:

• That the employer primarily relied on an ADS to make the decision;
• That a human reviewed the decision and corroborated the ADS output;
• How to reach a person for more information, and that you have a right to a description of your own data the system used; and
• That the employer is prohibited from retaliating against you for exercising these rights.

Access to your data. Under Section 1522(d), when an employer has primarily used an ADS to discipline or fire you, you can request, and the employer must provide, a meaningful, objective description of your own data the system used. Information about customers or co-workers is anonymized. The law gives you a description of the data, not necessarily the software’s source code or every raw record, but it is often enough to spot errors, such as a missed punch that was really a broken scanner, or a “low rate” recorded while you were on an approved break or accommodation.

Who Is Covered, and the Gaps to Know About

The definition of “employer” is wide. It includes anyone who directly or indirectly, or through an agent, employs or controls the wages, hours, or working conditions of a worker. It also expressly includes state and local public employers, labor contractors, and farm labor contractors. The protections run to “employees,” meaning anyone the employer employs.

Several limits are worth knowing:

• Job applicants and hiring decisions. SB 947 does not expressly cover candidates or hiring and recruiting decisions. Applicants screened out by an algorithm are not left without protection, though. FEHA still prohibits discrimination in hiring, and California Civil Rights Council regulations that took effect in October 2025 confirm that FEHA applies when employers use automated-decision systems.
• Union contracts. The law does not apply to parties covered by a collective bargaining agreement that clearly and unambiguously waives it, expressly provides for wages and working conditions, and provides protection from algorithmic management.
• Certain federal work. There is a narrow exception where using an ADS is required by, or reasonably necessary to comply with, a federal statute, regulation, or binding federal contract involving aircraft development for the national airspace, or national security, military, space, or defense products and services.
• Local laws can go further. SB 947 does not preempt city or county ordinances that give workers equal or greater protection.

How SB 947 Is Enforced

The Labor Commissioner can investigate violations, order temporary relief to preserve the status quo, issue citations, and file civil actions. Public prosecutors can also enforce the law. Violations carry a civil penalty of $500 per violation, and the statute contemplates injunctive relief, punitive damages, and attorney’s fees and costs in civil actions brought under its enforcement section (Lab. Code § 1526.1).

Two features are especially helpful to workers:

• Anti-retaliation. Section 1526 bars an employer from firing, threatening, demoting, suspending, or otherwise retaliating against a worker for using or trying to use SB 947 rights, filing a Labor Commissioner complaint, alleging a violation, or cooperating in an investigation.
• The burden shifts to the employer. In an enforcement proceeding over the corroboration rules, once it is shown that an ADS was used to make the decision, the employer must prove either that it did not primarily rely on the ADS or that it followed the corroboration and notice requirements (Lab. Code § 1526.1(c)). Employers that keep poor records of their human review will have a hard time.

The statute’s enforcement provisions are written around the Labor Commissioner and public prosecutors. Whether an individual worker can sue directly under SB 947 is not spelled out, and courts will likely sort that out over time. In practice, though, an algorithm-driven firing usually raises other claims a worker can bring, which is where FEHA comes in.

How SB 947 Works With FEHA Discrimination and Retaliation Claims

SB 947 adds process rules. It does not replace the protections workers already have. FEHA makes it unlawful to discriminate in discipline, pay, promotion, or termination because of a protected characteristic, and to retaliate against someone for opposing discrimination or harassment, requesting an accommodation, or participating in an investigation. Here is how the two laws fit together:

• “The computer did it” is not a defense. An employer is responsible for the tools it chooses. If an algorithm’s scores fall harder on older workers, workers with disabilities, pregnant workers, or a particular racial or national-origin group, that can support a FEHA discrimination claim, whether the bias was intended or built into the data.
• Proxies can be discrimination. Systems that infer protected traits are banned under SB 947. Even without an explicit inference, factors like zip code, gaps in attendance tied to medical leave, accent-scoring in call centers, or “speed” metrics that ignore an accommodation can act as stand-ins for protected traits.
• Retaliation by algorithm is still retaliation. If a system penalized you for time off you were legally entitled to, for a complaint you made, or for supporting a co-worker’s claim, that may violate FEHA’s anti-retaliation rule, CFRA leave protections, and SB 947’s ban on using an ADS to punish workers for exercising legal rights. See our guide to FEHA retaliation in California.
• A missing notice or skipped review is evidence. If your employer primarily relied on software but never gave you the required notice, or no one actually checked the output, that gap can help show the stated reason for your firing was not the real reason. Pretext is often the heart of a discrimination or retaliation case. Our article on how to prove wrongful termination walks through how that evidence is used.

FEHA claims usually start with a complaint to the California Civil Rights Department (CRD), which generally must be filed within three years of the adverse action. Remedies can include lost wages, emotional distress damages, punitive damages in appropriate cases, and attorney’s fees. Learn more about workplace discrimination claims.

What to Do If You Were Disciplined or Fired by an Algorithm

1. Ask directly whether software was involved. Ask, in writing, what information the decision was based on and whether a scoring, tracking, or AI system played a role. Keep the answer.
2. Save every notice and screenshot. Keep the postuse notice if you received one, plus any automated warnings, scores, dashboards, app messages, and emails. If you were supposed to get a notice and didn’t, write that down along with the date you were told about the decision.
3. Request a description of your data. For decisions on or after July 1, 2027, where the employer primarily relied on an ADS, send a written request for the description of your own data the system used. Date it and keep a copy.
4. Request your personnel file and payroll records. California workers and former workers have separate rights to inspect or copy their personnel files (Lab. Code § 1198.5) and payroll records. Here is how to request your personnel file.
5. Write your own timeline. Note when you complained, requested leave or an accommodation, reported a safety issue, or disclosed a medical condition, and when the warnings or scores started. Patterns matter. Our tips on how to document workplace retaliation can help.
6. Look for errors in the data. Equipment failures, approved breaks, accommodations, schedule changes, and mismatched shifts are common reasons an automated score is wrong.
7. Be careful with severance and separation paperwork. Don’t sign a release until you understand what claims you may be giving up.
8. Act quickly. Deadlines vary by claim, and some Labor Commissioner retaliation complaints must be filed within one year. Talking to a lawyer early protects your options.

What About Discipline Before July 1, 2027?

SB 947’s ADS rules are not operative until July 1, 2027, so they generally won’t apply to a write-up or firing that happens before then. That does not mean workers are unprotected in the meantime. FEHA, CFRA, California’s whistleblower laws, wage-and-hour retaliation protections, and the Civil Rights Council’s automated-decision regulations already apply. If an algorithm was used to push you out because of a protected trait or because you spoke up, you may have a claim today.

Practical Takeaways

• Beginning July 1, 2027, software cannot be the sole basis for discipline or termination in California.
• When software is the primary basis, a human must corroborate it with real information, and the worker must get a written notice.
• Workers can request a description of the data the system used about them.
• Algorithms may not be used to infer protected traits or to target workers for exercising legal rights.
• Applicants and hiring decisions fall outside SB 947, but FEHA still applies to automated hiring tools.
• An algorithm-driven firing can support a FEHA discrimination, retaliation, or wrongful-termination claim, and SB 947 violations can be strong evidence.

If you were written up, demoted, or fired and you suspect a scoring system, tracking app, or AI tool was behind it, especially after you complained, took leave, or asked for an accommodation, contact Jonny Law for a free consultation. We can help you figure out what happened, what records to request, and which claims may be available. This post is general information about California employment law as of November 12, 2026. It is not legal advice about any specific case.

Sources: SB 947 (McNerney), Stats. 2026, ch. 859, adding Labor Code §§ 1520–1526.7; Fenwick, “California Gov. Gavin Newsom Signs New Employment Bills into Law for 2027 and Beyond”; National Law Review, California employment bill signing summary. Related: 2027 California employment law updates.