Lawsuit Alleges Meta Used AI to Target Employees on Protected Leave for Mass Layoffs

Twenty-six current and former Meta employees filed a federal lawsuit in Oakland, California, alleging the social media giant used a "constellation" of artificial intelligence tools to select workers for mass layoffs. The plaintiffs claim the AI systems disproportionately targeted employees who were on legally protected medical, parental, or disability leave.[1][3]

The layoffs in question stem from Meta's May 2026 workforce reduction, which eliminated approximately 8,000 jobs—roughly 10% of the company's total headcount. The cuts were part of a broader corporate restructuring aimed at transforming Meta into an "AI-first" organization. However, the lawsuit argues that the very technology Meta is pivoting toward was weaponized against its own workforce during the termination process.[3][5]

At the center of the complaint is the mechanism by which Meta allegedly evaluated employee performance. According to the 71-page filing, the company relied on a suite of internal AI systems to score and rank staff. These included "Metamate," an internal large language model assistant, as well as employee-trained "second-brain" agents designed to replicate parts of a worker's output.[5][6]

The plaintiffs allege these AI tools were fed by extensive employee-monitoring data. The lawsuit claims the systems tracked keystrokes, screen content, mouse activity, browser history, and AI token-usage dashboards to generate algorithmic performance rankings. This data was then allegedly used to assemble the final termination list without adequate human oversight.[1][6]

The core legal claim hinges on how these algorithmic systems handled periods of inactivity. The plaintiffs argue that because the AI relied heavily on continuous output metrics and software development activity, it structurally disadvantaged anyone who took time away from work. Employees on approved leave naturally generated fewer keystrokes, sent fewer messages, and consumed fewer AI tokens.[4][5]

By design, the lawsuit alleges, the AI systems failed to account for these gaps. Instead of pausing the metrics or conducting a "leave-neutral review," the algorithms effectively recorded medical leave, maternity leave, and disability accommodations as severe underperformance. One plaintiff, a scientist, was reportedly selected for redundancy just two days before giving birth, while on approved pre-birth maternity leave.[2][7]

The plaintiffs, who filed anonymously, span six states and the District of Columbia. They all share one commonality: each took, requested, or was approved for protected leave within the past 24 months. Their claims include violations of the Family and Medical Leave Act (FMLA), the Americans with Disabilities Act (ADA), the Pregnancy Discrimination Act, and various state-level worker protections.[4][6]

Meta has strongly pushed back against the allegations. A company spokesperson dismissed the claims as lacking merit, stating unequivocally that "workforce management and organizational decisions were and are made by people, not AI." The company maintains that human managers, not algorithms, finalized the layoff selections and that protected leave status was not a factor in those decisions.[2][5]

The defense highlights a critical tension in modern enterprise AI: the distinction between an algorithm providing data and an algorithm making a decision. Meta's position suggests that while AI tools may have been used to track productivity or assist in performance reviews, the ultimate responsibility for the termination list rested with human judgment.[5]

However, the plaintiffs contend that human managers were largely bypassed. The complaint alleges that "Meta did not assemble the termination list through the considered judgment of managers who knew the work," but instead deferred to the AI-generated rankings. In one cited instance, an employee's manager allegedly warned him that submitting a leave application "will definitely" result in termination, suggesting the system's outcomes were known to be punitive.[1][7]

The lawsuit also points to Meta's broader performance calibration system, known internally as Checkpoint. Introduced earlier this year, Checkpoint reportedly placed a heavy emphasis on outcomes and made employee AI adoption a "core assessment metric." Workers were categorized into tiers such as "AI Native" or "AI Enabled," with scores declining when they were away from their keyboards.[5]

This case represents what legal observers believe is the first major lawsuit against a U.S. technology giant challenging the use of artificial intelligence in conducting mass layoffs. It arrives as regulators increasingly scrutinize automated decision systems in the workplace. States like California, Colorado, and Illinois have recently passed or proposed laws aimed at protecting workers from algorithmic bias.[1][6]

The plaintiffs are seeking a preliminary injunction from the court to block Meta from finalizing the layoffs, which are scheduled to take effect on July 22. While the workers were notified in May, they technically remain employed by the company until the separation date. They are asking to maintain the status quo while they pursue their claims in private arbitration.[2][4]

Beyond halting the terminations, the lawsuit demands an independent audit of Meta's AI tools. The plaintiffs' legal team argues that Meta has deliberately kept the mechanics of its selection process secret, and an audit is necessary to reveal exactly how the algorithms weighted periods of protected leave.[1][5]

The outcome of this dispute could set a significant precedent for how enterprises deploy AI in human resources. As companies increasingly adopt AI-driven productivity tracking and performance management software, the legal requirement to ensure these systems do not inadvertently violate established labor laws—such as the FMLA and ADA—will become a central compliance challenge.[1][6]