[Into the World of AI] AI Firing Humans... From the Meta Lawsuit to the ‘No Robo Bosses Act’

California Enacts Law Banning AI-Only Dismissals
Mandatory Human Oversight for AI-Based HR Decisions
Former Meta Employees File Class-Action Suit Alleging "Unfair Dismissals by AI"

The so-called 'Robo-boss' practice, in which artificial intelligence (AI) evaluates worker productivity and autonomously decides on dismissals, is facing simultaneous pushback from the U.S. judiciary and local governments. As California becomes the first state in the U.S. to enact a law prohibiting layoffs solely by AI, a wave of wrongful dismissal lawsuits from former employees against big tech companies is also intensifying.

According to CNBC and other sources on October 1 (local time), California Governor Gavin Newsom signed the 'No Robo Bosses Act' (SB 947) into law on September 30. This legislation restricts AI from making unilateral decisions regarding employee termination or discipline. It is the first instance in the U.S. where human oversight and review of algorithmic human resources management is legally mandated.

With this law taking effect, companies in California are now required to have a human manager review any decisions to dismiss or discipline employees made by automated decision systems (ADS), using performance records and any other additional data. Also, if an AI system is used as a key tool in human resource decisions, employers must formally notify the employee in writing and explain the data behind the decision. Furthermore, the installation of AI surveillance devices in employee break rooms, restrooms, or other facilities, as well as the use of tools that collect biometric data to infer workers’ emotional states, is completely prohibited. The law will take effect on July 1 next year.

Separate from these government-level legislative regulations, the industry is already witnessing legal battles over malfunctions of AI-powered HR systems. Twenty-six former employees of Meta, the parent company of Facebook and Instagram, have filed a class-action wrongful termination lawsuit against the company in the U.S. District Court for the Northern District of California.

Meta logo. Reuters·Yonhap News

According to the plaintiffs’ complaint, Meta used its internal AI system to collect employees’ keyboard activity, email volumes, and work activity data in real time to assign productivity scores. The problem was that the system failed to recognize gaps caused by legally protected sick leave or parental leave during performance evaluations. In one case, a researcher participating in the lawsuit was terminated while on company-approved parental leave, and another manager was included in a round of layoffs just two weeks after starting sick leave.

The plaintiffs argued, "When computer usage data stopped accumulating during leave, the AI system simply classified these employees as underperformers and automatically marked them for termination." In response, Meta denied all allegations, stating in a written reply that "all workforce reductions and organizational decisions were made by human managers according to neutral criteria, not by AI."

According to a survey by the Organisation for Economic Co-operation and Development (OECD) last year, 90% of U.S. companies already use algorithmic management tools to assign, monitor, or evaluate work tasks for employees. Tech companies have promoted AI-driven HR tools as delivering fairer assessments by eliminating human subjectivity and bias.

However, both Meta’s class-action lawsuit and California’s new legislation serve as clear examples of how data-driven automated evaluations can actually infringe upon rights protected by labor law. CNBC commented, "The Meta lawsuit starkly illustrates concerns about the impact of AI on the disabled community in the workplace and labor market." On California’s move, they noted, "This could spur other states to pursue workplace AI-related legislation."

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