Twenty-six former employees of Meta Platforms have launched legal action against the technology company in federal court in Oakland, California, asserting that algorithmic systems were deployed to systematically disadvantage workers with disabilities or those requiring medical leave during recent rounds of redundancies. The lawsuit, filed anonymously late on a Monday in July, represents the latest challenge to Meta's aggressive restructuring efforts and raises broader questions about how artificial intelligence is being deployed in human resources decisions across Silicon Valley and beyond.
The core allegation centres on Meta's reliance on quantitative performance metrics that disadvantaged employees unable to maintain full productivity due to medical circumstances. According to the legal filing, the company weighted factors including productivity measurements and AI token usage—a metric tracking computational resource consumption—when determining which staff members to eliminate. For workers who had taken legitimate medical leave or who worked at reduced capacity due to health conditions, these metrics would inherently reflect absences rather than capability or value, creating a systematic bias in the selection process.
The timing of the lawsuit comes months after Meta announced in early 2024 that it would eliminate approximately 10 per cent of its global workforce, representing nearly 8,000 positions. The company subsequently announced additional rounds of job cuts to follow. These reductions represented one of the largest workforce contractions in Meta's history and significantly exceeded typical attrition levels, suggesting that systematic processes were required to identify candidates across the entire organisation.
The 26 plaintiffs, who have chosen to remain anonymous in the filings, represent employees spread across six American states including California and New York, as well as the District of Columbia. By pursuing their claim under both federal and state employment law, the group is drawing on multiple legal frameworks that explicitly prohibit discrimination against workers with disabilities, those utilising protected medical leave, and pregnant employees. The geographic diversity of the claimants suggests the alleged practices were not isolated to particular offices or divisions but rather reflected centralised decision-making at the corporate level.
For Malaysian readers and observers across Southeast Asia, this case illuminates a critical governance gap in how multinational technology companies apply artificial intelligence to employment decisions. As regional tech firms increasingly adopt automated systems for workforce management, the Meta lawsuit serves as a cautionary example of how algorithmic systems can encode and amplify discrimination, even when designers do not explicitly intend discriminatory outcomes. The metrics selected for measurement—productivity, resource consumption—were inherently neutral on their face, yet their application created disparate impact on a protected class of workers.
Meta's response to the allegations has been characterised by categorical denial rather than engagement with the underlying methodology concerns. The company's spokesperson stated that workforce decisions were made by people rather than artificial intelligence, a position that sidesteps the fundamental allegation. The lawsuit does not claim that AI alone made decisions; rather, it contends that AI systems were used to narrow the pool of candidates considered for layoffs, thereby exercising decisive influence over which employees faced termination. This distinction between AI making decisions autonomously and AI substantially shaping the decision-making process remains a crucial point of contention in employment law.
The legal landscape surrounding algorithmic discrimination in hiring and workforce management is still evolving, particularly in the United States. Several states have begun implementing requirements for bias testing of employment algorithms, and the Equal Employment Opportunity Commission has increasingly scrutinised automated decision-making systems. However, much of this regulatory development remains reactive—responding to alleged harms rather than establishing proactive standards. The Meta case may accelerate such regulatory development by demonstrating how productivity metrics embedded in AI systems can systematically disadvantage protected groups.
From a practical standpoint, the allegations raise questions about how organisations should structure mass layoff decisions when significant portions of their workforce have legitimately taken medical leave or work accommodations. If productivity metrics are used without adjustment or context regarding the circumstances producing those metrics, they inevitably become a proxy for disability status or medical leave usage. Organisations seeking to defend against discrimination claims would need to demonstrate that their selection criteria were directly related to legitimate business needs and that protected characteristics played no material role in the decisions.
The case also reflects broader tensions within the technology industry regarding the speed and scale of workforce reductions. When a company needs to eliminate thousands of positions rapidly, human resources teams typically lack the capacity to conduct individual assessments of each affected employee. This operational constraint creates pressure to rely on automated systems that can process large volumes of data quickly. However, this efficiency comes at the cost of losing individualised context that might reveal how particular metrics should be interpreted for employees with legitimate reasons for lower productivity figures.
Beyond the immediate legal implications for Meta, the lawsuit signals to other large employers considering algorithmic approaches to workforce decisions that disability discrimination risks are substantial and cognisable in court. For Southeast Asian companies expanding into technology services and considering adoption of similar systems, the case provides an important reminder that algorithms are not neutral tools. They encode the values and assumptions of their designers, and when applied to employment contexts, they can produce outcomes that violate principles of fair treatment that most jurisdictions now recognise as fundamental to employment law.
The discovery process in this case is likely to produce significant documentation regarding how Meta designed its selection algorithms, what data was included, how metrics were weighted, and whether managers raised concerns about potential discrimination. Such documentation could become material not only to this specific case but also to regulatory scrutiny by government agencies and to public understanding of how these systems operate. The case may ultimately establish important precedents regarding the responsibility of employers who use AI systems in high-stakes employment decisions.
