Overview
- Twenty-six current and former Meta employees filed a federal lawsuit alleging a network of internal AI and monitoring tools flagged workers who took medical, parental, or disability leave for termination.
- The complaint names systems such as a large language model assistant called Metamate, an employee 'second brain,' and productivity measures drawn from keystrokes, emails and browser activity as factors that fed layoff decisions.
- U.S. District Judge William Orrick denied the plaintiffs' emergency request to stop the layoffs, saying the workers have not yet produced direct evidence showing how AI influenced Meta's choices.
- A preliminary injunction hearing is set for August 24, and Orrick said he could revisit relief if plaintiffs uncover evidence about improper AI use that rebuts Meta's claim that humans made the final decisions.
- The case highlights barriers to proving algorithmic discrimination because key data is controlled by the employer and many workers are bound to confidential arbitration, which limits public discovery and testing of workplace AI.