A group of 26 former Meta Platforms employees has launched legal action against the social media corporation, contending that the company weaponised artificial intelligence systems to disproportionately identify and dismiss workers with disabilities or those who had taken medical leave. The lawsuit, submitted to federal court in Oakland, California on Monday, represents a significant challenge to Meta's approach during its recent restructuring exercise and raises troubling questions about how algorithmic tools are deployed in employment decisions.
According to the legal complaint, Meta relied on AI-powered metrics centred on productivity measures and artificial intelligence token usage when determining which employees should be eliminated from payroll as part of the company's broader workforce reduction strategy. This methodology, the plaintiffs argue, created a systematic bias against individuals whose work patterns were disrupted by health-related absences—a group that includes employees managing chronic illnesses, recovering from medical procedures, or dealing with disabilities that occasionally required them to step back from their duties. The consequence was that workers already facing workplace disadvantages became even more vulnerable during the layoff process.
Meta's restructuring began in May this year when the company announced plans to eliminate approximately 10 percent of its global workforce, translating to nearly 8,000 positions. The redundancies were executed in waves throughout the year, creating widespread uncertainty across the organisation and the broader technology sector. For workers with existing health challenges, these layoffs compounded their vulnerability, as the selection criteria appeared to penalise those whose medical circumstances had affected their work output metrics.
The 26 plaintiffs, who have chosen to proceed anonymously through the court filing, are not limited to residents of a single state. Their geographic diversity—spanning California, New York, and Washington DC, plus three additional states—suggests the problem extended across Meta's operations rather than being confined to any single regional office. This geographic spread also indicates that the algorithmic selection process operated consistently across multiple Meta facilities and departments, reinforcing the claim that the discrimination was systematic rather than incidental.
The legal action specifically charges Meta with breaching both federal statutes and state legislation designed to protect employees from discrimination based on disability status, retaliation for using medical leave, and pregnancy-related conditions. These laws, which exist in most developed nations and are similarly enshrined in Malaysian and broader Southeast Asian employment frameworks, establish that employers cannot make adverse employment decisions based on protected characteristics or circumstances. The plaintiffs contend that Meta's use of AI in the selection process circumvented the protective intent of these laws by obscuring the discriminatory logic beneath layers of algorithmic complexity.
Meta's official response to the allegations came swiftly, with company representatives asserting that the claims are without foundation. The company's statement emphasised that "workforce management and organisational decisions were and are made by people, not AI," suggesting that human judgment remained paramount in the layoff selections. This defence raises the critical question of whether and to what extent humans actually scrutinised the AI recommendations before making final determinations, or whether the algorithmic output was largely accepted without meaningful oversight.
The lawsuit touches on a growing anxiety within the global technology industry regarding the transparency and fairness of AI deployment in employment contexts. As companies increasingly adopt machine learning systems to streamline hiring, promotion, and termination decisions, concerns mount that these tools may perpetuate or amplify existing biases embedded in historical data. If algorithms are trained on patterns reflecting past employment decisions, they may replicate historical discrimination against protected groups.
For Malaysia and the Southeast Asian region, this case carries significant implications. As regional technology companies expand their operations and increasingly adopt AI-driven management tools, employers face a critical choice: either implement robust safeguards ensuring algorithmic fairness, or risk similar legal exposure. Malaysian employment law, particularly provisions within the Employment Act and the Industrial Relations Act, similarly prohibits discrimination based on disability and medical conditions. A Meta defeat could establish important precedent suggesting that companies cannot shield themselves from discrimination liability by delegating decisions to algorithms.
The case also highlights tensions between efficiency gains and ethical employment practices. While AI systems can theoretically improve objectivity by removing human bias, they can simultaneously introduce new forms of discrimination if not carefully designed and monitored. Meta's layoff approach apparently prioritised metrics like productivity and token usage without accounting for legitimate variations in performance resulting from protected medical circumstances—a fundamental oversight that legal frameworks in Malaysia and internationally are designed to prevent.
The lawsuit's outcome could shape how multinational technology firms, including those with significant operations in Southeast Asia, approach workforce reductions and personnel decisions going forward. Should the plaintiffs prevail, companies may face pressure to adopt more transparent, auditable AI systems with explicit safeguards for vulnerable employee populations. The alternative—continued reliance on opaque algorithms without meaningful human oversight—increasingly appears legally and ethically indefensible in jurisdictions with robust employment protections.
As employment law globally grapples with the implications of artificial intelligence, this case underscores that technology cannot replace human judgment in consequential decisions affecting workers' livelihoods. The broader question now facing the technology industry and regulatory bodies across the Asia-Pacific region is whether sufficient safeguards exist to prevent algorithmic systems from becoming instruments of discrimination, however unintentionally.
