Meta sued by 26 former employees who allege AI tools unfairly targeted workers on leave for layoffs

When AI Picks Who Gets Fired: The Meta Lawsuit That Could Rewrite the Rules of Algorithmic Employment

Twenty-six former Meta employees are walking into a courtroom with a question that every company using AI in HR should be terrified to answer: when an algorithm ranks workers for termination and a human says "approved," who actually made the decision?

The lawsuit, filed against Meta and first reported by Reuters, alleges that the company used a "constellation" of internal AI tools to score, rank, and select employees for layoffs. The 26 plaintiffs say those tools systematically penalized workers who were on medical and parental leave 1.

The layoffs happened in May, part of Meta's plan to cut roughly 10 percent of its workforce, or about 8,000 people 1. What makes this case different from every other tech layoff story is the mechanism. The plaintiffs allege that Meta deployed Metamate (its internal AI assistant), employee-trained AI agents, and dashboards tracking AI token usage to generate performance rankings. Those rankings, the suit claims, became the termination list 1.

Here is the core allegation: the AI tools failed to account for protected leave. Workers who were out on medical or parental leave had lower activity metrics, less AI tool engagement, and fewer visible contributions during their absence. The algorithm, apparently blind to the reason for that absence, scored them lower. The lawsuit says this "in effect penalized the employees for exercising their legal rights to these leaves" 1.

Meta's response is a single sentence that perfectly captures the legal gray zone at the heart of this case. Spokesperson Tracy Clayton told The Verge: "These claims lack merit and are not based on facts. Workforce management and organizational decisions were and are made by people, not AI" 1.

Think about that defense for a second. A manager walks into a room with a ranked list generated by an AI system. The list has already sorted employees by score. The manager reviews the bottom 10 percent and signs off. Did a human make the decision? Technically, yes. But was the decision shaped, constrained, and effectively pre-determined by an algorithm? Also yes.

That tension is what makes this lawsuit genuinely unprecedented. Employment discrimination law in the United States was built around human decision-makers. The framework asks whether a manager had discriminatory intent, whether they followed fair procedures, whether they can justify their reasoning. It was never designed for a world where the reasoning happens inside a model and the human's role is reduced to a rubber stamp.

The lawsuit accuses Meta of violating both federal and state laws that protect employees from being terminated for taking protected leave 1. If the plaintiffs prevail, the logic extends well beyond Meta. Any company that uses AI tools to inform performance reviews, reduction-in-force decisions, or even promotion pipelines would need to confront the same question: can you prove your algorithm did not discriminate?

So what?

If you build AI tools for HR or workforce management, this case draws a line around your liability. The lawsuit alleges that Meta's tools scored, ranked, and selected employees without accounting for protected status 1. That is a design failure, not just a legal one. If your product feeds employment decisions, it needs to be auditable, explainable, and explicitly designed to exclude protected categories from its inputs. Build for the day someone subpoenas your model's training data, because that day is coming.

If you invest in AI-in-HR companies, the unit economics of that category may be about to shift. An adverse ruling or settlement could impose compliance, audit, and documentation costs that reshape margins across the sector. The companies best positioned are those already building explainability and bias-testing into their products. The ones most exposed are those selling "AI-powered" tools with no audit trail.

If you are an employee, this case could determine whether you have legal recourse when an algorithm gets your evaluation wrong. The plaintiffs allege they were penalized for taking legally protected leave because the AI could not distinguish between "low performer" and "on medical leave" 1. If that happened at Meta, it is almost certainly happening elsewhere. The question is whether the law will catch up to the technology.

The Meta lawsuit is not really about Meta. It is about the moment we are in, where AI tools are quietly embedded in the most consequential decisions a company makes about its people. Meta says humans made the calls. The plaintiffs say the algorithm made the calls and humans just signed the paperwork. Whoever the court believes will set a precedent that reaches every workplace in America.

The signal: when a human rubber-stamps an algorithm, "a human approved it" may no longer be a defense.