California is one step away from passing one of the nation’s most consequential workplace AI laws.
Senate Bill 7, known as the “No Robo Bosses” Act, has cleared the California State Assembly and Senate and now awaits Governor Gavin Newsom’s signature. If enacted, the bill would impose sweeping new obligations on employers that use automated decision systems, defined broadly to include machine learning tools, algorithms, and artificial intelligence, to make employment-related decisions. These obligations would extend across the employment lifecycle, covering areas such as hiring, performance management, promotion, scheduling, termination, and even productivity tracking.
SB 7 Arrives at a Pivotal Moment
California’s Civil Rights Council recently finalized regulations on automated decision systems in employment, which take effect on October 1, 2025. Unlike the Council’s rules, which focus on discriminatory outcomes, SB 7 targets procedural fairness and transparency. Together, they represent a layered and complementary compliance regime that will reshape how employers deploy technology in the workplace.
Governor Newsom’s office has not signaled a position on SB 7. The governor has previously vetoed other AI-related legislation, citing premature regulatory frameworks and the need for further industry consultation. However, the political and public pressure surrounding the use of artificial intelligence in employment decisions is growing, and this bill may prove more difficult to dismiss. If signed, SB 7 would take effect on January 1, 2026.
What SB 7 Would Require
At its core, SB 7 is about visibility and accountability. The bill requires employers to notify workers and job applicants when an automated decision system is in use and to ensure that high-impact decisions are not made by machines alone.
Employers would be required to give written notice to any worker who is likely to be affected by the use of an automated decision system, excluding hiring-related systems, at least thirty days before deployment. For systems already in use when the law takes effect, employers would have until April 1, 2026, to provide notice. New hires must receive notice within thirty days of onboarding.
The bill also mandates transparency in hiring. If an employer plans to use an automated decision system in evaluating a candidate for a specific role, that information must be disclosed at the point of application. This can be communicated through an automatic email response or embedded directly into the job posting.
Each notice must describe the categories of employment decisions affected, the types of worker data collected, the sources of that data, and how the system processes it. Employers must also disclose whether the system measures productivity quotas, how quotas may change, and what adverse actions may result from noncompliance.
Perhaps most notably, SB 7 prohibits employers from relying solely on an automated decision system to discipline, terminate, or deactivate a worker. If an employer uses such a system to guide these decisions, the employer must incorporate human review and consider other forms of evidence, including personnel files, supervisory evaluations, and peer feedback. The bill emphasizes that customer ratings may not be used as the primary input for automated decisions, a provision that may significantly impact gig economy platforms and customer-facing industries.
In addition, workers would have the right to request a copy of the personal data used by an automated decision system in decisions involving discipline, termination, or deactivation. Employers would be required to provide that data once per year and must ensure that the data is anonymized to protect the identities of other individuals, such as customers or coworkers.
Enforcement and Exceptions
SB 7 includes multiple enforcement mechanisms. The Labor Commissioner would have the authority to investigate complaints and issue citations. Public prosecutors could also bring civil actions, and workers could seek injunctive relief, punitive damages, and attorney’s fees through private lawsuits. Each violation could carry a civil penalty of $500.
The bill includes limited carve-outs. Employers that are parties to valid collective bargaining agreements may be exempt from SB 7’s requirements, provided the agreement explicitly waives coverage, includes provisions addressing wages and working conditions, and protects workers from algorithmic management. The bill also recognizes that some employers may be required to use automated decision systems under federal law or as a condition of federal contracts. In those instances, employers may continue to meet federal requirements without violating the law.
While SB 7 limits overlapping state requirements related to automated decision system notices, it expressly preserves local ordinances that offer greater protections. This means that employers in jurisdictions like San Francisco or Los Angeles may be subject to additional compliance burdens.
A New Layer: Civil Rights Council Regulations
SB 7 does not exist in a vacuum. It arrives just as the California Civil Rights Council’s final regulations on automated decision systems are set to take effect on October 1, 2025.
The Council’s rules operate under the authority of the Fair Employment and Housing Act (FEHA) and prohibit the use of automated decision systems that result in discrimination based on protected characteristics, such as race, gender, disability, age, or national origin. The rules emphasize the employer’s responsibility to prevent adverse impact and to offer reasonable accommodations when automated tools may disadvantage certain applicants or employees.
For example, an algorithm that filters resumes based on gaps in employment could be found discriminatory if it disproportionately excludes individuals with disabilities or women returning from caregiving leave. Employers using automated video interviews that assess facial expressions or vocal tone must be especially cautious, as those tools could run afoul of FEHA’s protections against national origin, disability, and gender bias.
The Council also imposes strict recordkeeping requirements. Employers must retain all data related to the use of automated decision systems, including applications, test results, scoring data, and impact analyses, for at least four years. These records must be made available to the Civil Rights Department upon request.
Unlike SB 7, the Council’s regulations do not require notice to workers, do not create a right of access to data, and do not impose a prohibition on fully automated decision-making. Instead, they focus on outcome-based accountability and are primarily enforced through administrative complaints.
A Combined Compliance Challenge
If SB 7 becomes law, employers will need to navigate both sets of rules. The overlap between the two frameworks is substantial, but their approaches are distinct. SB 7 addresses how employers use automated decision systems. The Civil Rights Council focuses on whether those systems produce discriminatory effects.
Together, they create a two-pronged compliance mandate. Employers will need to ensure that workers are informed about automated tools and that decisions are not made in a black box. At the same time, employers must continuously test and monitor these systems to ensure they do not create disparate impacts on protected groups.
What Employers Should Do Now
Regardless of whether Governor Newsom signs the bill, employers should begin preparing. First, organizations should inventory all automated decision systems currently in use, including scheduling software, productivity monitors, and AI-enhanced screening tools. That inventory should include the purpose of each system, the data it relies on, and the vendor that developed it.
Next, employers should draft plain-language notices that comply with SB 7’s requirements. These notices should be reviewed by counsel to ensure alignment with FEHA’s anti-discrimination mandates and recordkeeping obligations.
Employers should also establish protocols for human oversight, particularly for decisions that carry legal or reputational risk. These protocols should require human reviewers to consider other forms of evidence and should be documented as part of each employment decision. Finally, companies should train managers, HR professionals, and compliance teams on both sets of rules. A well-informed workforce is an employer’s first line of defense against regulatory enforcement and litigation.
Parting Thoughts
California continues to lead the nation in shaping the rules of the road for AI in employment. Whether SB 7 is signed or vetoed, its advancement through the legislature signals a growing demand for transparency, fairness, and human judgment in algorithmic management. Employers that take proactive steps now will be better positioned to adapt to whatever comes next, not just in California, but across the country.
The bots may not be taking over, but the regulators certainly are.