A new federal lawsuit claims Meta didn’t just use AI to trim its workforce — it used algorithms that systematically discriminated against employees on parental, medical, or family leave. The complaint, filed in California, argues the company’s internal AI systems generated the layoff lists that led to 8,000 job cuts in May, and that those systems disproportionately flagged workers in protected leave categories. One plaintiff got her pink slip a mere 48 hours before her due date.
The suit alleges the selection criteria baked into these systems included performance ratings, productivity metrics, and even how frequently staff used AI tools. The implication is clear: if you were out on leave — and therefore not generating code or logging AI interactions — the algorithm may have seen you as less productive. It’s the kind of brutal, data-driven logic that crunches numbers without grasping legal or human context.
Meta’s response is blunt. A company spokesperson flatly denied the allegations, insisting that “humans make all personnel decisions.” That’s a clean, defensible line, and it’s worth remembering that algorithmic recommendations are almost never the final word in large organizations. But the existence of AI-generated candidate lists, if proven, raises uncomfortable questions about how much rubber-stamping happened after the machines did their sorting.
The plaintiffs aren’t just seeking damages — they want a preliminary injunction to halt their terminations while the arbitration plays out. That’s a high bar, and it suggests they believe the paper trail is on their side. Whether this reveals a flaw in Meta’s AI governance or simply a misunderstanding of how internal tooling works, it’s a stark preview of litigation we’ll see more of as workforce algorithms move from hiring to firing.
