26 Ex-Meta Staff Sue Over AI-Driven Layoff Bias

What happened

Twenty-six former Meta employees have filed a lawsuit in federal court in Oakland, California, alleging that the company's AI-driven layoff process discriminated against workers who were on protected leave. The suit was filed late Monday, ahead of a July 22 date when employee separations are set to begin.

The layoffs in question stem from Meta's broader plan announced in May to cut roughly 8,000 jobs — about 10% of its workforce.

The core allegation

According to the lawsuit, Meta used what plaintiffs describe as a "constellation" of internal AI tools to score, rank, and select employees for termination. These reportedly included:

Metamate (Meta's internal AI assistant)
Employee-trained AI agents
Keystroke and activity-monitoring data
AI token-usage dashboards

The plaintiffs claim this system did not exclude or adjust for employees on parental, medical, or other legally protected leave. Because scores or ratings allegedly could not be accumulated by employees who were out on leave — or whose output was reduced due to disability — the suit argues Meta should have paused the system for an individualized, leave-neutral review, but did not.

All 26 plaintiffs reportedly had taken protected leave or requested/received a disability accommodation:

8 plaintiffs are women who took maternity or pregnancy-related leave
4 plaintiffs are men who took parental leave
1 plaintiff is a woman who took leave for family caregiving and later bereavement leave

The lawsuit cites alleged violations of the Family and Medical Leave Act (FMLA), the Americans with Disabilities Act (ADA), the Pregnancy Discrimination Act (PDA), and the Pregnant Workers Fairness Act (PWFA).

Meta's response

Meta spokesperson Tracy Clayton pushed back on the allegations, stating the claims "lack merit and are not based on facts," and adding that "workforce management and organizational decisions were and are made by people, not AI."

What's still unclear

The report notes several open questions: it's not clear how the May layoffs relate to the July 22 separation date — whether they're the same round or a distinct phase. There's also no detail on how the AI scoring algorithms actually worked, what data fed into termination decisions, or what internal review process (if any) Meta conducted before the layoffs. The timeline for resolving the lawsuit and the specific remedies plaintiffs are seeking remain undisclosed.

Why founders should care

This case is likely to be closely watched as an early test of how courts treat AI-assisted HR decisions under existing leave and disability laws. For founders building or buying workforce-management tools, a few implications seem probable:

Startups that rely on activity data, token usage, or algorithmic scoring for performance reviews may face elevated legal risk if those systems don't explicitly exclude or adjust for employees on protected leave.
Purely automated ranking systems without human oversight could become a reputational liability, especially as more public disputes like this one surface.
Companies that build in safeguards — such as pausing automated scoring for employees on leave, or requiring human review before AI-influenced terminations — may be better positioned to limit exposure under FMLA, ADA, PDA, and PWFA-type statutes.
Documentation of how AI tools are used in personnel decisions could increasingly matter for legal defensibility, particularly as regulators and courts scrutinize these practices more closely.

The opportunity