Twenty-six Meta employees filed a lawsuit on July 13, 2026 in the US District Court for the Northern District of California, alleging that AI algorithms — not managers — made the decisions about who would lose their jobs. Among the 26 plaintiffs are people who were on maternity or medical leave at the time they were flagged for termination. According to Reuters, this is reportedly the first lawsuit of its kind filed against a major US tech company.
Key takeaways
- 26 anonymous (Doe) plaintiffs filed the lawsuit on July 13, 2026 in federal court in California
- Meta used a system called Metamate and AI agents to score employees before the layoff wave
- Workers were classified by AI adoption level: AI Native?AI Native: Meta's internal performance tier for employees with the highest level of AI tool usage. Higher adoption scores correlated directly with better standing in the algorithmic ranking system., AI First, AI Enabled
- The layoffs target roughly 10% of staff — first separations scheduled for July 22, 2026
- Meta denies wrongdoing: decisions were made by people, not AI
How the employee scoring system worked
According to the complaint, Meta did not ask managers to independently evaluate their reports. Instead, the company relied on a constellation of internal AI systems — including a tool called Metamate, AI agents trained on employee data called second-brain agents, dashboards tracking each employee AI token consumption, and an algorithmically assisted performance ranking system.
The central metric was a measure of how much each employee used Meta internal AI tools. Workers were assigned to one of three categories: AI Native (highest AI usage), AI First, or AI Enabled. People who could not use these tools intensively — for example because they were on maternity or medical leave — could not accumulate points in this system by design.
Those tools draw on inputs — performance ratings, calibration scores, productivity and output metrics, AI-native ratings, and AI-token consumption — that, by design, cannot be accumulated by an employee who is on protected medical or family leave.
— from the lawsuit complaint
Pregnancy and leave on the termination list
The lawsuit includes detailed accounts of individual cases. One plaintiff — described as a scientist — was selected for termination while on approved pre-birth pregnancy leave, the day before her water broke and two days before she gave birth.
Other plaintiffs were on maternity or paternity leave, and some were working under approved remote-work accommodations for disabilities. Their legal team argues that the scoring system made no adjustments for any of these circumstances.
Context worth noting: Meta announced these layoffs in May 2026 despite reporting record revenue the previous month. The company simultaneously committed to AI spending of $125-145 billion for 2026 — more than double its 2025 budget. Employees publicly questioned why layoffs were necessary when the company was performing at record levels. The lawsuit cites this contrast directly.
Plaintiffs demand audit and employment freeze
The 26 plaintiffs are seeking a preliminary injunction blocking the finalization of layoffs until an independent audit of the selection process is conducted. The proposed audit would examine the AI system inputs and weights, determine whether protected-leave or disability status was used as an input or effectively penalized, and recalculate each plaintiff's score with appropriate adjustments.
The plaintiffs invoke the US Family and Medical Leave Act, the Pregnancy Discrimination Act, the Americans with Disabilities Act, the Pregnant Workers Fairness Act, and various state laws — including California's updated Fair Employment and Housing Act, which explicitly prohibits the use of automated
The lawsuit is not a class action — Meta requires employees to sign arbitration agreements that waive the right to participate in class actions. The plaintiffs are pursuing individual arbitrations but argue that a court order is necessary to preserve their employment status during that process.
Meta, in a statement provided to Ars Technica, said: These claims lack merit and are not based on facts. Workforce management and organizational decisions were and are made by people, not AI.
Why this matters
This lawsuit opens a new front in the debate about algorithmic HR management. Until now, most AI-in-HR discussions have focused on hiring — the problem of discriminatory candidate screening algorithms is relatively well documented. Using AI to generate layoff lists is a different scenario with a harder-to-trace discrimination mechanism.
Measuring an employee AI-native score as a proxy for productivity is particularly problematic. Someone on maternity leave literally cannot use internal tools, so their score drops — not because they are less productive, but because they are exercising a legally protected right. If a court or arbitrator rules this constitutes systemic discrimination, the entire tech industry may need to fundamentally rethink how productivity algorithms are applied in personnel decisions.
The broader question is who bears responsibility when a human uses an AI-generated list. Meta's assurance that people made the decisions may not be sufficient if those people simply rubber-stamped algorithmic outputs without independent review.
What's next
- A court must rule on the preliminary injunction request before Meta carries out separations scheduled for July 22, 2026
- Regardless of the injunction outcome, 26 cases will proceed to individual arbitrations under Meta employment agreements
- If an audit is ordered, its findings could set a precedent for regulating AI use in HR processes across the tech industry
