Twenty-six Meta employees say the company’s own artificial intelligence turned protected medical and parental leave into a layoff target list.
Story Snapshot
- Workers claim Meta used AI scoring to pick thousands of layoffs, hitting people on medical and family leave hardest.
- The lawsuit says tools tracked keystrokes, screen content, emails, browser history, and AI use, punishing anyone who stepped away.
- Employees on maternity, parental, disability, and caregiving leave say the system never paused or adjusted for their protected absences.
- Meta denies the claims, insists humans made the decisions, and faces a landmark test of AI and workplace rights.
A federal lawsuit puts Meta’s layoff engine under a microscope
Twenty-six current and former Meta employees filed a federal lawsuit in Oakland, California, accusing the company of using artificial intelligence systems that disproportionately targeted workers on medical, parental, or family leave for layoffs.
They are part of a larger group of about 8,000 employees, roughly 10% of Meta’s workforce, that Meta said it would lay off as part of a major restructuring. These plaintiffs say the way Meta chose who stayed and who left was not neutral, fair, or legal.
The complaint says Meta did not rely on traditional performance reviews or individual meetings when it selected workers for termination. Instead, it allegedly turned to internal AI-assisted systems and activity tracking tools to score and rank employees for a termination list.
The lawsuit points to keystroke logging, screen content scanning, email and browser history tracking, and dashboards showing how often workers used Meta’s own AI tools. Those numbers became the yardstick for who looked “productive” enough to keep.
How AI scoring can turn protected leave into a liability
The workers’ core argument is simple and sharp: you cannot rack up high productivity scores when you are home with a newborn, in surgery, or caring for a sick parent.
The lawsuit says many of Meta’s scores and 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”.
If the system measures only visible activity, then anyone who legally steps away from a keyboard falls behind by default. That is where the alleged discrimination begins.
According to the complaint, Meta did not pause these systems or adjust the scoring rules when employees took protected leave. The workers say the company “did not pause the system for the individualized, leave- and accommodation-neutral review that the law requires”.
In plain terms, they claim Meta let an uncorrected algorithm treat time spent on a hospital bed or in a nursery as laziness or low output. For Americans who value family, faith, and personal responsibility, that picture collides hard with fairness.
The human stories behind the numbers
About half of the plaintiffs had taken leave for caregiving or pregnancy-related reasons. The group includes eight women who had taken maternity or pregnancy-related leave, four men who had taken parental leave, and one woman who took time to care for a family member and later bereavement leave. Another plaintiff reported disability-related limits on work and says Meta offered no adjustment for her condition.
One account described a layoff notice landing while a worker was on approved pre-birth leave just days before giving birth. These stories put faces and families behind the charts and code.
The lawsuit argues that people on protected medical or family leave were disproportionately selected for layoffs compared with peers who stayed continuously at their desks. That is the heart of the legal claim.
If true, it means a system that was supposed to help management make data-driven decisions instead turned legally protected choices into hidden risk factors.
For older readers who watched factories and offices go through wave after wave of downsizing, this feels like the latest version of the same old fear: the tools change, but the people getting squeezed often look familiar.
The laws the workers say Meta broke
The plaintiffs accuse Meta of violating a long list of federal and state protections. The lawsuit cites the Family and Medical Leave Act, which guards workers who take approved medical or family leave from punishment.
It invokes the Americans with Disabilities Act, which bars discrimination and demands reasonable accommodations for people with disabilities.
It also points to pregnancy-related protections, including the Pregnancy Discrimination Act and the Pregnant Workers Fairness Act, which seek to keep expectant and new mothers from being pushed out of work. Combined, these laws reflect a clear national value: you should not have to choose between your job and your health or family.
According to reporting, the lawsuit also claims Meta failed to test its AI systems for bias, despite new requirements in places like California and New York City. Those rules grew out of a concern that algorithms can quietly bake in unfairness if nobody checks them.
This is straightforward: if the government sets guardrails, big companies are expected to follow them, not treat them as suggestions while they chase efficiency.
Meta’s denial and the proof problem
Meta has publicly pushed back hard, calling the claims “not based on facts” and saying “workforce management and organizational decisions were and are made by people, not AI”. The company insists that humans, not software, made the layoff calls. That sets up a clear clash.
The workers say these human decisions leaned heavily on AI-driven rankings and scores. Meta says AI did not drive the outcome. Resolving that dispute will depend on emails, dashboards, ranking sheets, and testimony that the public has not yet seen.
The 26 workers asked a federal judge to block Meta from completing the layoffs while they pursue individual discrimination claims in private arbitration.
Early coverage notes that a judge has been cautious about halting the process, stressing how hard it is for workers to prove what happened when they “were not in the rooms where it happened” and the company controls the data. That highlights a larger problem with algorithmic management.
When decisions rest on opaque scoring systems, ordinary people may feel the impact in their bank accounts long before they ever see the logic that sealed their fate.
Sources:
abc7.com, theguardian.com, reuters.com, youtube.com
