California SB 947

Overview

California SB 947 adds Part 5.5.5, commencing with Section 1520, to Division 2 of the Labor Code and creates a workplace framework governing automated decision systems (ADS) used by employers. The statute is chaptered as Chapter 859, Statutes of 2026, and, although approved and filed on 2026-09-30, it becomes operative on 2027-07-01. It applies to all cities, including charter cities, and is severable. The law focuses on employment decisions that materially affect workers, especially discipline and termination, and prohibits employers from using ADS in ways that would evade labor, occupational safety, employment, or civil rights law, infer protected status, or punish workers for exercising legal rights.

The statute also establishes notice, human-review, and access rights when an employer primarily relies on ADS output to discipline or terminate an employee. In that situation, the employee may request a meaningful, objective description of the employee’s own data used by the ADS, and the employer must provide a stand-alone, written post-use notice in plain language. Enforcement is assigned to the Labor Commissioner and also to public prosecutors. The law authorizes citations, civil actions, temporary relief, injunctive relief, punitive damages, attorney’s fees and costs, and a civil penalty of $500 per violation. It also contains exceptions for certain collective bargaining agreements, for operations required by or reasonably necessary to comply with federal law or binding federal contracts in narrow aerospace, national security, military, space, or defense contexts, and for privacy-related ADS regulations adopted under California privacy law.

Definitions

SB 947 defines “automated decision system” broadly 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, and is used to assist or replace human discretionary decisionmaking while materially impacting natural persons. The statute separately defines “ADS output” to include information, data, assumptions, predictions, scoring, recommendations, decisions, or conclusions generated by an ADS. It also defines “artificial intelligence” in functional terms as an engineered or machine-based system that can infer from inputs how to generate outputs that influence physical or virtual environments.

The law defines employer coverage expansively. “Employer” includes persons who directly or indirectly control wages, hours, or working conditions, and also includes all branches of state government, cities, counties, special districts, transit districts, UC upon agreement by the regents, CSU, community college districts, school districts, and other governmental entities. It also covers labor contractors and farm labor contractors. “Employee data” is defined broadly as any information that identifies, relates to, or describes an employee, regardless of how it is collected, inferred, or obtained. The statute expressly excludes ordinary tools and infrastructure such as spam filters, firewalls, antivirus software, identity and access management tools, calculators, databases, datasets, and other compilations of data.

Governance and Institutional Framework

Primary enforcement authority rests with the California Labor Commissioner, who may investigate alleged violations, order appropriate temporary relief to preserve the status quo, issue citations, and file civil actions. The statute expressly links enforcement procedures to existing Labor Code mechanisms, including Sections 98.3, 98.7, 98.74, and 1197.1, which means the law is designed to operate through familiar wage-and-hour style administrative and civil enforcement channels. Public prosecutors may also bring civil enforcement actions under Chapter 8 of Division 1, broadening the institutional capacity to respond to unlawful ADS use in the workplace.

The statute also creates a burden-shifting structure in litigation and administrative proceedings. Once a plaintiff shows that an ADS was used to make a disciplinary or deactivation decision, the employer must demonstrate either that it did not primarily rely on the ADS or that it complied with the corroboration and notice requirements. This structure is important governance-wise because it makes ADS use a trigger for stronger employer accountability rather than a neutral technological fact. The law also preserves stronger local protections by stating that it does not preempt any city, county, or city and county ordinance providing equal or greater employee protection, while simultaneously limiting duplicative notice obligations only where they are substantially similar and not otherwise preserved by privacy regulations or specific exceptions.

Key Focus Areas

The central policy objective is to prevent algorithmic discipline and termination from becoming fully automated or opaque. Employers may not use ADS to violate labor, occupational safety, employment, or civil rights law, infer a worker’s protected status under Government Code Section 12940, or predict and take adverse action against a worker for exercising legal rights. The statute therefore targets discriminatory, retaliatory, and legally evasive uses of workplace analytics. It is not a general ban on AI in employment; instead, it is a targeted restriction on high-impact uses that affect individual workers’ livelihoods.

A second focus area is human oversight. Employers may not rely solely on ADS when making a disciplinary or termination decision. If an employer primarily relies on ADS output, it must direct a human to corroborate the decision using the data used to produce the output or other corroborating or supporting information, such as supervisory evaluations, personnel files, work product, peer reviews, and witness interviews, including relevant online customer reviews. If the output cannot be corroborated or the reviewer concludes it is inaccurate, incomplete, or misleading, the employer may not use it for discipline or termination. The statute also centers transparency by requiring a notice that explains the ADS role in the decision and by giving employees a right to a meaningful description of the data used about them.

Implementation Framework

Implementation begins on 2027-07-01, when the operative provisions take effect. At that point, employers must ensure that workplace ADS use is screened against the statute’s prohibitions and that disciplinary or termination processes involving primary reliance on ADS include a human corroboration step. The law does not require employers to abandon automated tools entirely; rather, it requires them to structure processes so that humans review and substantively corroborate critical decisions before the ADS output is used to harm an employee. This means employers will need internal procedures for documenting the basis of the human review and preserving the corroborating information used in that review.

For notice and employee-access rights, the employer must provide a written post-use notice at the time it informs the employee of the disciplinary or termination decision. The notice must be a separate, stand-alone communication, written in plain language, provided in the language used for routine employee communications, and delivered through a simple written method such as email, a hyperlink, or another written format. When the employee requests the data description, the employer must provide a meaningful, objective description of the employee’s own data used by the ADS, and any customer, co-worker, or other personal information in that description must be anonymized. Employers covered by a valid collective bargaining agreement can be excluded only if the agreement explicitly waives the statute in clear and unambiguous terms, covers wages and working conditions, and provides protection from algorithmic management.

Monitoring and Evaluation

SB 947 does not create a dedicated agency auditing program, but it embeds monitoring through notice, documentary access, and enforcement. The requirement to provide a post-use notice creates a traceable event whenever an ADS primarily drives a disciplinary or termination decision. The employee’s right to request a description of the employee’s own data used by the ADS creates a second layer of review, enabling workers to challenge or understand the basis for the decision. The statute’s requirement that the human reviewer corroborate the ADS output with specific types of supporting information is also a practical monitoring mechanism because it forces comparison between machine output and conventional workplace evidence.

For ex post evaluation, the enforcement system is significant. The Labor Commissioner may investigate, issue citations, and seek relief, and a public prosecutor may bring civil actions. In litigation, the employer bears the burden after ADS use is shown. That burden shift is designed to make compliance assessable rather than merely aspirational. The law also allows temporary injunctive relief, which can stop harmful practices while a matter is pending. Together these features create a compliance architecture that is evidence-driven and reviewable after the fact, even though the statute does not assign a separate technical auditing body or prescribe formal impact-assessment templates.

Penalties, Liability, and Appeals

SB 947 includes both anti-retaliation protections and civil enforcement remedies. Employers may not discharge, threaten to discharge, demote, suspend, or otherwise discriminate or retaliate against an employee for using or attempting to use statutory rights, filing a complaint with the Labor Commissioner, alleging a violation, cooperating in an investigation or prosecution, or otherwise assisting enforcement. This makes retaliation itself an independent violation, separate from the underlying improper ADS use.

On liability and remedies, the statute authorizes the petitioner in a civil action to seek appropriate temporary or preliminary injunctive relief, punitive damages, and reasonable attorney’s fees and costs, in addition to other remedies available under the section. The statute also sets a civil penalty of $500 per violation. It does not create a bespoke appellate pathway in the text; rather, disputes proceed through the normal administrative and civil processes associated with Labor Commissioner citations and civil actions. The statute further preserves local ordinances offering equal or greater employee protection and prevents displacement of certain privacy-related automated decisionmaking regulations applicable to businesses subject to the California Consumer Privacy Act of 2018.

Relationship to Other Instruments

The statute expressly interacts with California privacy law by providing that businesses subject to the California Consumer Privacy Act of 2018 remain subject to any privacy-related automated decisionmaking technology regulation adopted by the California Privacy Protection Agency under Civil Code Sections 1798.185 and 1798.199.40. It also states that an employer complying with the statute’s notice requirements need not comply with any substantially similar notice provisions under any other state law, except as otherwise specified. That anti-duplication rule is limited and does not extend to quotas under Labor Code Section 2100 or other automated standards applied to working conditions.

SB 947 also coexists with collective bargaining and federal requirements. It is inapplicable where a valid collective bargaining agreement explicitly waives the part in clear and unambiguous terms and includes the statutory bargaining subject matter and protections from algorithmic management. In addition, the statute does not prohibit ADS use where required by, or reasonably necessary to comply with, a federal statute, federal regulation, or binding federal contract for aircraft development in the national airspace or for national security, military, space, or defense purposes. The measure is also framed as a matter of statewide concern applying to charter cities, which places it within California’s state-local legal hierarchy rather than treating workplace algorithm rules as a purely municipal matter.

National/Federal Alignment

At the federal level, SB 947 does not implement a specific federal AI statute; instead, it creates a state labor standard that sits alongside federal employment, labor, civil rights, and privacy law. The statute’s design is consistent with federal anti-discrimination and labor-protection principles because it prevents ADS from being used to evade legal obligations or to punish workers for asserting legal rights. At the same time, it leaves room for employers who must comply with binding federal law or contracts in sensitive aerospace and national-security contexts, reflecting a narrow accommodation of federal operational requirements.

The statute differs from federal law by imposing a direct state-law rule on workplace ADS use, a human-corroboration requirement, and a mandatory post-use notice right. It also explicitly preserves stronger local protections and certain California privacy regulations, making it a layered compliance regime rather than a single national standard. No federal counterpart is identified in the statute text itself. Practically, employers operating in California must therefore harmonize SB 947 with federal civil rights and labor obligations, while treating the California requirements as an additional state layer for any ADS-driven discipline or termination decision.

Implementation Timeline

Compliance Checklist

Sources and References