News · 2026-07-17

26 Meta Workers Sue, Saying AI Picked Them for Layoffs While on Leave

A group of 26 Meta employees sued the company in federal court in Oakland, California on July 14, 2026, alleging that Meta used AI systems to select workers for layoffs in a way that disproportionately targeted people on medical, parental and family leave. It is one of the first major cases in which a big tech company's own internal AI selection process -- not bias in an AI product it sells, but AI used to decide who gets fired -- is the central legal allegation.

Key facts

• All 26 plaintiffs took protected leave and requested or received a disability accommodation; they are among the roughly 8,000 employees (about 10% of staff) Meta said it would cut in May, with separations set to begin July 22.
• The suit alleges Meta used 'internal AI systems, keystroke and activity-monitoring data, AI token-usage dashboards and algorithmically assisted performance rankings' to pick layoffs.
• Meta calls the claims meritless; reporting is from the Associated Press via ABC News, CBS News and Fortune.

The core of the complaint is a mechanism, not just a grievance. The plaintiffs argue that many of Meta's performance scores and activity ratings, 'by design, cannot be accumulated by an employee who is on protected medical or family leave, or whose output is reduced by a disability.' When you are out on maternity leave or recovering from a serious health condition, you generate fewer keystrokes, fewer tokens, fewer commits -- and if an algorithm reads that quiet period as low performance, it systematically penalizes exactly the people the law protects. The plaintiff group is composed accordingly: eight women who took maternity or pregnancy leave, four men who took parental leave, one woman who took family-care and bereavement leave, and others with approved disability accommodations. One plaintiff says a manager warned him that taking approved medical leave would get him picked for the layoff.

The legal theory is 'disparate impact' -- the doctrine that a facially neutral policy can still be illegal if it disproportionately burdens a protected class. The lawyers argue Meta's 'algorithmically assisted selection process, by systematically recording such absences as reduced performance, falls more heavily on women than on men.' The suit invokes the Family and Medical Leave Act, the Americans with Disabilities Act, the Pregnancy Discrimination Act and the Pregnant Workers Fairness Act. Notably, this private suit proceeds even as the current administration has ordered federal agencies to deprioritize disparate-impact enforcement -- workers can still bring their own claims, and several state laws prohibit disparate-impact discrimination outright.

Meta's defense is a single sentence with a lot riding on it: 'Workforce management and organizational decisions were and are made by people, not AI.' That is the crux of the whole case. Meta is not disputing that AI tools and data dashboards existed; it is asserting that humans made the decisions, with AI merely informing them. The plaintiffs counter that the score-generation itself was structurally biased, so a human rubber-stamping a biased score does not launder the bias. Whether 'a person clicked the button' is enough to escape liability when the ranking feeding that button was algorithmic is the question courts will have to answer.

Why it matters: as companies wire AI and activity-monitoring into performance management, this case tests whether AI-driven HR decisions generate the same anti-discrimination liability as human ones -- and whether 'the model made me do it' works as a defense. It rhymes with a quieter theme running through the week's developer-fatigue essay: work flowing into AI systems that then judge the humans who produced it, with no one on the other side accountable for how the scoring works.

The honest caveat: these are allegations, not adjudicated findings. Meta denies them, no court has ruled, and the separations have not yet taken effect. But with a July 22 layoff date looming over 26 people who remain employed as they sue, the case is both timely and unusually concrete about the specific data -- token dashboards, keystroke logs, algorithmic rankings -- that plaintiffs say decided their jobs.

Primary source, verified: read the paper →

Ground Truth. (2026, July 17). 26 Meta Workers Sue, Saying AI Picked Them for Layoffs While on Leave. Ground Truth. https://groundtruth.day/news/meta-employees-sue-ai-layoff-selection.html

@misc{groundtruth:meta-employees-sue-ai-layoff-selection,
  title  = {26 Meta Workers Sue, Saying AI Picked Them for Layoffs While on Leave},
  author = {{Ground Truth}},
  year   = {2026},
  month  = {jul},
  url    = {https://groundtruth.day/news/meta-employees-sue-ai-layoff-selection.html}
}

Topics: policy · labor · meta · ai-in-employment · legal