26 Meta employees just sued over AI-driven layoffs. This is the first case of its kind. Every company using AI for workforce decisions is now on notice.

Meta used a constellation of AI systems to pick who gets fired.

"Metamate" LLMs. Employee-trained "second brain" agents. Keystroke monitoring. AI token usage dashboards. Algorithmically assisted performance rankings.

The lawsuit says these systems scored, ranked, and selected 8,000 people for termination in May.

Here is what the algorithms missed.

People on medical leave cannot generate keystrokes. Parents on parental leave cannot consume AI tokens. Disabled workers with reduced output cannot compete on productivity dashboards.

By design, these metrics penalize absence. The lawsuit says Meta never adjusted for protected leave. Never excluded leave-takers from the selection cohort. Never paused for the individualized review the law requires.

26 employees are still employed. Separations start July 22.

Meta's response: "Workforce management decisions were made by people, not AI."

That defense just got tested in federal court. The lawsuit invokes FMLA, ADA, the Pregnancy Discrimination Act, and the Pregnant Workers Fairness Act. It also invokes disparate impact liability, the same civil rights doctrine the Trump administration tried to kill.

If your organization uses any AI system to evaluate, rank, or select employees for termination, promotion, or restructuring, audit it today.

The metrics your system tracks may be illegal to use. The outputs it generates may constitute discrimination. And the "people made the decision" defense may not hold when the algorithm built the list.

This case sets the template. More are coming.

TITLE: Meta used AI to fire people. Now 26 employees say the algorithm was biased against disabled workers and parents.

BODY:

Twenty-six Meta employees just filed the first lawsuit in history challenging AI-driven workforce decisions.

Filed Monday night in Oakland federal court, the complaint lays out how Meta built a termination list using AI systems that, by design, penalized people who weren't at their desks.

The tools: "Metamate" LLMs, employee-trained "second brain" agents, keystroke and screen monitoring, AI token usage dashboards, algorithmically assisted performance rankings.

The result: people on medical leave, parental leave, and disability accommodations were disproportionately selected for the 8,000-person layoff announced in May.

One employee was discouraged from taking approved medical leave by a manager who warned it would trigger layoff selection. Another received the first lower rating of his career in a cycle covering protected leave. Eight women who took maternity leave are among the plaintiffs.

Meta's defense: decisions were made by "people, not AI."

But the lawsuit argues the system never adjusted for protected leave. Never excluded leave-takers from the selection pool. Never paused for the individualized review federal law requires.

Separations start July 22. The plaintiffs want a preliminary injunction to block them.

If your company uses AI tools to evaluate, rank, or score employees, this case is your warning.

Audit your metrics. The signals your system tracks may be illegal to use. And "the algorithm decided" is not a defense. It is an exhibit.