A federal judge in California has cleared the way for Meta Platforms to proceed with layoffs affecting 26 workers who contend the company deployed artificial intelligence systems to identify and terminate employees with disabilities or those who took medical leave. U.S. District Judge William Orrick in Oakland declined to issue an emergency order halting the terminations scheduled to begin on July 22, determining that the plaintiffs had not demonstrated the "irreparable harm" necessary to justify such intervention while the case proceeds through private arbitration.
The lawsuit, filed anonymously by engineers, managers, researchers and designers, represents what legal observers regard as the first major challenge against a prominent American technology company over the alleged use of AI in conducting workforce reductions. The 26 workers allege that Meta systematically relied on artificial intelligence tools—including a large language model assistant dubbed "Metamate" and a system that tracked employee communications and documents—to score and rank staff for termination in a manner that disadvantaged those with health conditions or caregiving responsibilities.
According to court filings, the company's approach to selecting which positions to eliminate incorporated metrics derived from productivity scoring systems that monitored employee keystroke activity, screen content, emails and browsing patterns. The plaintiffs further contend that Meta factored AI adoption scores into the layoff selection process, but these algorithms did not account for periods when workers were on vacation or legally protected medical leave, causing their engagement metrics to decline artificially and placing them at greater risk of termination.
In May, Meta announced it would eliminate approximately 8,000 positions, representing roughly 10 percent of its global workforce, as the company intensifies its strategic pivot toward artificial intelligence development and deployment. The targeted workers claim this reduction in force was not primarily driven by human judgment, but rather by algorithmic systems that systematically disadvantaged specific employee cohorts based on protected characteristics or circumstances.
Judge Orrick's Friday ruling addressed only the plaintiffs' request for a temporary restraining order to block the immediate layoffs. The judge suggested he remained open to reconsidering his position once presented with additional evidence about how and whether the company deployed AI tools in its reduction in force decisions. The workers' broader motion for a preliminary injunction, a longer-lasting temporary protective order, remains pending before the court, leaving room for potential intervention if new information emerges about Meta's processes.
The plaintiffs' legal representatives emphasized during Thursday's hearing that the stakes extended beyond mere loss of income. Departing workers face the prospect of forfeiting valuable stock options and employer-provided health insurance coverage, potentially imperiling access to medical care for pregnancies, ongoing treatments, and other health conditions. Attorney Barbara Cowan highlighted the irreversible consequences: "There's no do-over for bonding with a new baby or giving birth or having active medical treatment." Meta's counsel countered that affected workers would retain the ability to obtain health coverage independently, characterizing the losses as typical employment-related damages that could be recovered later if the plaintiffs prevail in arbitration.
A central procedural issue complicating the case involves Meta's mandatory arbitration agreements, which generally require employees to resolve workplace disputes individually rather than pursuing collective action in court. However, the plaintiffs argue these agreements contain carve-outs permitting employees to seek temporary relief in judicial proceedings—a legal exception the workers invoke to justify their request for emergency intervention. While exceptions allowing temporary relief requests are not uncommon in arbitration agreements, they traditionally apply to scenarios involving alleged theft of proprietary information or poaching of clients and staff, rather than layoffs of at-will employees, making this application relatively novel.
Meta has consistently denied any wrongdoing, insisting that humans made the ultimate decisions regarding the layoffs and that no discrimination occurred. The company's position reflects a broader industry tendency to emphasize human oversight and decision-making authority even when acknowledging the use of algorithmic tools in preliminary screening or evaluation stages. According to court filings, affected workers remained on the payroll following the May notifications but lost access to Meta's computer systems on May 20 and have performed no work for the company since that date.
For Malaysian and Southeast Asian observers, this case carries significant implications as regional technology companies increasingly adopt similar AI-driven workforce management systems. The lawsuit raises fundamental questions about algorithmic transparency, fairness, and accountability in the employment relationship—issues that extend beyond individual companies to shape regulatory approaches across multiple jurisdictions. As nations in the region grapple with how to govern artificial intelligence deployment, the American legal system's treatment of employment discrimination claims rooted in algorithmic decision-making may establish important precedents influencing both corporate practices and regulatory frameworks.
The dispute also underscores growing tension within the technology sector between productivity-focused metrics and employee protections. Meta's emphasis on AI adoption and productivity scoring reflects a management philosophy that privileges measurable algorithmic output, yet this approach creates particular vulnerability for workers experiencing temporary health challenges, disability accommodations, or family caregiving demands. The court's indication that it may revisit its determinations suggests awareness among American judges that AI-driven employment decisions warrant heightened scrutiny and may demand different evidentiary standards than traditional human-centered layoff processes.
As the case progresses through arbitration, the broader question of whether artificial intelligence systems can legally and fairly conduct workforce reductions remains largely unanswered. The judge's openness to further evidence on "whether and how AI was used" in the reduction in force indicates recognition that current legal frameworks may be inadequate to address technological developments outpacing established employment law. This foundational uncertainty will likely shape how technology companies approach similar challenges in coming years, particularly as AI systems become more autonomous and their decision-making processes more opaque.
The outcome of this litigation could influence how companies across Southeast Asia design and deploy workforce optimization systems, particularly in jurisdictions with stronger employee protections than those in some American employment contexts. Whether courts will ultimately hold employers accountable for discriminatory outcomes produced by algorithmic systems—even when human review theoretically remains part of the process—may determine whether artificial intelligence becomes a tool for more objective hiring and termination decisions or simply a mechanism for obscuring and automating existing workplace biases.
