26 Meta employees sue over alleged AI‑driven layoff picks

Twenty-six current and former Meta employees have filed a federal lawsuit alleging the company used internal AI systems, keystroke and activity monitoring, and algorithmically assisted performance scores to select workers for May reductions — a process the plaintiffs say disproportionately targeted employees on medical, parental or family leave.

The suits, filed July 13–14, 2026, claim the selection practices violated the Family and Medical Leave Act, the Americans with Disabilities Act and the Pregnancy Discrimination Act, and amount to disparate‑impact discrimination against workers who took protected leave. Plaintiffs say internal tools and metrics flagged people while they were on leave or shortly after return, and that those signals were used to recommend cuts to human decision‑makers.

Meta has pushed back, saying the May reductions were determined by people rather than by autonomous algorithms. The company maintains that managers made layoff decisions and that any internal scores or monitoring tools were only advisory. Plaintiffs counter that even advisory algorithmic outputs and continuous monitoring can steer human choices and create unlawful disparities when they correlate with protected characteristics such as pregnancy or disability.

The complaint pinpoints several features of Meta’s internal workforce systems: keystroke and activity logging that measures responsiveness and daily patterns, scores that combine engagement and performance indicators, and automated ranking or filtering workflows that plaintiffs say fed into layoff candidate lists. According to the filings, workers on short‑term medical leave, parental leave and intermittent family‑care leave were overrepresented among those ultimately selected for separation.

Employment lawyers said the suit tests a growing legal question: when does algorithmic assistance cross from informational to determinative, and what statutory protections apply when the outputs of workplace surveillance tools correlate with protected leave status? "Courts will look at whether the tools had a disparate impact and whether the employer can show a business necessity that could not be met by less discriminatory alternatives," said one employment lawyer who reviewed the filing.

The case lands amid a broader wave of scrutiny over algorithmic personnel decisions and workplace surveillance. Regulators and courts in the U.S. and Europe have stepped up inquiries into whether automated or algorithm‑assisted tools produce unlawful bias, while civil‑liberties groups have challenged keystroke monitoring and continuous productivity tracking as invasive and error‑prone. Employers have increasingly adopted these technologies to manage distributed workforces, but that adoption has raised fresh questions about how to comply with leave, disability and pregnancy protections.

What the filings do not show is a detailed map of Meta’s internal models or the specific thresholds and weights used to generate the contested scores. Plaintiffs allege a causal link between the tools and the outcome, but the company has not publicly disclosed the algorithms, training data, internal validation results or whether independent bias audits were conducted. Meta also has not provided a public account of whether safeguards — for example, human‑in‑the‑loop checks designed to detect leave‑related distortions — were in place and how they were applied during the May reductions.

The litigation could sharpen legal standards for employers that combine surveillance data and algorithmic scores in workforce decisions. If courts accept plaintiffs’ disparate‑impact claims, companies that rely on continuous monitoring or automated ranking in restructuring may face stiffer burdens to demonstrate nondiscriminatory necessity and to document bias mitigation. For HR leaders, the case underscores the operational and legal risk of tying layoff workflows to monitored productivity signals without transparent audits and clear safeguards for protected leave and disability status.

The plaintiffs seek classwide relief and damages; how the court treats discovery requests for Meta’s internal systems will likely determine whether the litigation produces public precedents on the use of AI and monitoring in layoffs. That outcome could influence not just tech companies but any employer weighing automated tools for performance assessments and redundancy planning.