After 17 Tardies, AI Says, "We Should Fire This One, Right?" ... Outcry as U.S. Puts Brakes on 'Robot Bosses'

California Bans Employee Dismissals Based Solely on AI Decisions

Controversy over accountability for employment decisions is growing in the United States amid allegations that artificial intelligence (AI) has monitored employees' work behavior and recommended dismissals, and was even used to select targets for large-scale layoffs. California has put legal safeguards in place to prevent employees from being fired or disciplined solely on the basis of AI decisions.

According to the Associated Press and other outlets, California Governor Gavin Newsom signed the No Robo Bosses Act (SB 947) on September 30, requiring human review during employee dismissal and disciplinary proceedings. It is the first law in the United States to require human oversight when automated decision-making systems, including AI, make decisions that affect employees' livelihoods.

Image to help readers understand the article. Getty Images Bank.

Even if AI selects employees for dismissal, people must verify the grounds

The law prohibits companies from firing or disciplining employees solely on the basis of the results produced by automated decision-making systems. If a system's judgment is used as a primary basis for a decision, a person must review relevant data, including managers' evaluations, personnel records, and employees' work output, to verify the grounds for the decision. If the grounds cannot be verified, or if the system's results are deemed inaccurate, incomplete, or misleading, those results cannot be used in a dismissal or disciplinary decision.

When notifying employees of a dismissal or disciplinary action, companies must inform them in writing that an automated system was used as a primary basis for the decision and that a person reviewed the system's results and verified the grounds for the decision. Employees may request an explanation of the data about them that the system used to make its decision. The regulations take effect on July 1, 2027.

AI has also recommended dismissing an employee for repeated tardiness

In the United States, calls for human review of employment decisions have grown as cases have emerged in which AI recommended dismissing employees. U.S. AI startup Andon Labs said on August 14 that Luna, its AI manager running Andon Market, an experimental store in San Francisco, had recommended dismissing an employee for repeated tardiness and violations of work instructions.

Andon Labs' tally of check-in messages showed that the employee had been late for 17 of 23 shifts. Luna recommended dismissal after considering the employee's repeated tardiness, failure to follow work instructions, and issues involving the management of a corporate credit card. Andon Labs said that after reviewing the recommendation, a person personally notified the employee of the dismissal.

Meta also faced controversy over a layoff process that used AI. In July, 26 employees filed a lawsuit alleging that the company used AI to select employees for dismissal and penalized workers who had taken parental or medical leave, among other types of leave. Meta denied the allegations, saying that employment decisions were made by people, not AI.

This content was produced with the assistance of AI translation services.