A California federal judge has allowed Meta Platforms to proceed with planned layoffs of 26 employees who allege the technology giant deployed artificial intelligence tools to systematically target workers with disabilities or those who took medical leave for job termination. The decision, handed down by U.S. District Judge William Orrick in Oakland on Friday, means the affected workers will lose their positions starting July 22, even as they pursue discrimination claims through private arbitration proceedings.
Orrick determined that the plaintiffs failed to demonstrate the "irreparable harm" threshold typically required for courts to issue emergency orders freezing corporate actions. Although the judge declined the workers' immediate relief request, his written ruling contained language suggesting he remains open to reconsidering the matter if additional evidence emerges about how Meta deployed artificial intelligence in the reduction process. This conditional position offers a potential opening for the plaintiffs, whose lawyers indicated they would continue pursuing a longer-term preliminary injunction.
The dispute centres on Meta's reliance on algorithmic systems during a dramatic workforce contraction announced in May, when the company notified approximately 8,000 employees—representing roughly ten percent of its global workforce—that their positions were being eliminated as Meta accelerated investments in artificial intelligence capabilities. The company has consistently maintained that human managers made all final decisions regarding terminations, though the plaintiffs present a markedly different picture of how technology influenced those determinations.
According to the lawsuit filed Monday, Meta's selection methodology incorporated multiple AI-assisted evaluation systems that inadvertently disadvantaged workers managing health challenges or family caregiving responsibilities. The plaintiffs specifically cite a large language model called "Metamate" that functioned as an employee-trained "second brain," continuously monitoring workers' communications and documents. The company also reportedly employed productivity scoring mechanisms that tracked keystrokes, screen content, emails, and browser history—metrics that naturally depressed scores for employees absent due to medical conditions or protected leave periods.
The case represents an unprecedented legal challenge against a major American corporation regarding algorithmic discrimination in mass layoffs. Legal experts have noted that while class action employment litigation is commonplace, claims specifically targeting AI-driven termination decisions break new ground. This distinction carries particular significance for Malaysia and Southeast Asian technology sectors, where similar algorithmic evaluation systems increasingly feature in multinational corporate operations. The outcome could establish important precedent regarding worker protections in jurisdictions with expanding AI adoption across human resources functions.
During Thursday's hearing, plaintiffs' attorney Barbara Cowan emphasized the cascading consequences extending beyond immediate income loss. Terminated workers face forfeiture of valuable stock options and employer-subsidized health insurance, with particular hardship for those managing ongoing medical conditions, pregnancies, or active treatment protocols. Cowan argued that certain losses—like opportunities to bond with newborns or continue critical medical care—cannot be remedied financially if plaintiffs ultimately prevail in arbitration months or years hence.
Meta's legal representative, Erin Connell, countered that workers retain access to health coverage through alternative channels, characterising the insurance loss as employer subsidies rather than coverage elimination. She further contended that traditional employment law remedies—including back pay and other monetary damages—adequately compensate for temporary job losses, establishing a conventional framework within which arbitration should function. This disagreement reflects broader tensions about whether employment law adequately addresses modern technological layoff mechanisms.
The workers' ability to seek emergency court relief despite mandatory arbitration agreements hinges on a technical distinction within contract law. While the plaintiffs' arbitration agreements require individual dispute resolution rather than collective class actions, most such agreements include exceptions for temporary restraining orders and injunctive relief—though these carve-outs historically apply to trade secret theft or employee raiding scenarios rather than at-will termination situations. The plaintiffs argue this exception extends to their circumstances, a position that could reshape how arbitration clauses function if successful.
The terminated employees—who filed suit anonymously—encompass diverse professional roles including engineers, managers, researchers, and designers. Following their May notification, workers lost system access on May 20 and ceased performing company functions, yet remained technically on payroll. Finalisation of their terminations is scheduled for July 22 for most workers, with additional separations continuing into late July and August. This staggered implementation timeline potentially allows judicial reconsideration between phases, though the judge's Friday decision suggests such intervention remains unlikely absent significant new evidence.
The underlying algorithmic systems Meta allegedly employed raise fundamental questions about how corporations should govern AI deployment in personnel decisions affecting livelihoods and health security. The company's failure to pause algorithmic evaluation systems during statutory leave periods—a basic safeguard acknowledged in the plaintiffs' complaint—suggests potential gaps between technical capability and responsible implementation practice. As Malaysian regulatory bodies increasingly scrutinise workplace AI applications, this American litigation provides valuable insight into algorithmic governance failures and judicial responses.
Orrick's language indicating willingness to reconsider demonstrates judicial awareness that initial emergency determinations may prove inadequate as factual records develop. The judge specifically invited parties to submit additional evidence regarding AI's actual role in reduction decisions, suggesting courts may yet prove willing to intervene if proof emerges that algorithmic systems functioned as primary rather than peripheral factors in termination selections. This judicial openness, combined with the novel legal questions at stake, ensures continued litigation momentum despite the immediate defeat.
For multinational technology companies operating across Southeast Asia, the case underscores emerging legal risks associated with algorithmic decision-making in human resources. Malaysian employment law, while not yet specifically addressing AI discrimination, provides potential frameworks through existing disability protections and medical leave provisions. Companies implementing similar productivity-tracking and algorithmic-ranking systems face increasing exposure to discrimination claims, particularly as worker awareness of algorithmic employment practices grows and legal systems develop coherent responses to technological change.
The proceedings ultimately illustrate the tension between corporate efficiency objectives and worker protection in technology-driven business transformation. Meta's investment acceleration in artificial intelligence created genuine operational pressures requiring workforce adjustments, yet the company's apparent failure to implement safeguards preventing algorithmic discrimination suggests preventable vulnerabilities. As artificial intelligence continues reshaping workplace management across Malaysia's growing tech sector, this litigation provides cautionary lessons about the regulatory gaps and human costs accompanying unconstrained algorithmic deployment.
