Meta faces a significant legal challenge after 26 of its employees filed a federal lawsuit in Oakland, California, alleging the company systematically used artificial intelligence tools to identify and target workers on protected medical, parental and family leave for redundancy. The complaint, lodged in mid-July, singles out the social media conglomerate's decision to deploy keystroke monitoring, activity tracking, AI token-usage dashboards and algorithmically-assisted performance rankings to determine which staff members to remove during its May announcement of an 8,000-person workforce reduction—roughly one-tenth of its total headcount at the time.
Central to the employees' case is the contention that Meta's AI-driven system was fundamentally incapable of fairly assessing workers whose output necessarily declined while they exercised their legal rights. The lawsuit emphasises that performance metrics and productivity scores "by design, cannot be accumulated by an employee who is on protected medical or family leave, or whose output is reduced by a disability." According to the filing, Meta did not recalibrate its automated selection protocols to account for protected absences, nor did it pause its systems to conduct the individualized, leave-neutral reviews that employment law requires. This structural flaw meant that employees exercising federally-protected rights—maternity leave, parental leave, medical leave for serious conditions—faced systematic disadvantage in the algorithmic scoring process.
The composition of the plaintiff group underscores the gendered dimensions of the alleged discrimination. Eight women had taken maternity or pregnancy-related leave, four men had taken parental leave, and one woman had taken leave to care for an ill family member before bereavement leave. Approximately half the group fell into caregiving or pregnancy-related leave categories. The lawsuit argues this pattern is not coincidental: because women disproportionately utilise pregnancy and family caregiving leave in the workplace, the company's failure to accommodate these absences in its performance measurements created a disparate impact that fell more heavily on female workers than their male counterparts.
One plaintiff's experience illustrates the alleged coercion embedded in Meta's approach. According to the lawsuit, an employee with an approved serious health condition and disability was actively discouraged from taking medical leave by management, who warned that doing so would make him vulnerable to selection for the anticipated redundancy. Meta offered no accommodation for his documented disability, the complaint states. This suggests the discriminatory impact extended beyond passive measurement failure into active deterrence—employees felt pressured to forgo their legal rights to avoid being marked for termination.
Meta has rejected the allegations, stating in a company statement that the claims "lack merit and are not based on facts" and asserting that "workforce management and organisational decisions were and are made by people, not AI." The company's position attempts to shift responsibility from its algorithmic systems to human decision-makers, a common defence in AI discrimination cases. However, this argument struggles to address the structural question at the heart of the lawsuit: regardless of who makes final decisions, if the data fed into those decisions systematically penalizes protected conduct, the outcome violates employment law.
The lawsuit invokes multiple federal and state legal frameworks, including the Family and Medical Leave Act, the Americans with Disabilities Act, the Pregnancy Discrimination Act and the Pregnant Workers Fairness Act. Significantly, the complaint also relies on the doctrine of "disparate impact"—a civil rights principle holding that facially neutral policies can constitute illegal discrimination if they disproportionately burden members of a protected class and lack job-related necessity. This legal theory has become contentious in the current political environment. The Trump administration has actively moved to deprioritise disparate impact enforcement, issuing directives to federal agencies arguing that the doctrine undermines meritocracy and encourages assumptions that workforce imbalances automatically reflect discrimination.
These political headwinds have already affected enforcement machinery. The Equal Employment Opportunity Commission has dropped discrimination complaints on behalf of some workers under the new administration's anti-disparate-impact stance. Yet the Meta lawsuit demonstrates that companies remain vulnerable to this form of litigation despite federal enforcement pullback, because workers retain the right to sue independently if the EEOC declines their complaints. Several states have also enshrined disparate impact protections in their own employment law, insulating state-level claims from federal retreat.
For Malaysian and Southeast Asian readers, this case carries broader implications about how technology companies deploy AI systems across hiring, performance evaluation and workforce reduction decisions. As regional tech sectors expand and mature, similar algorithmic systems are entering use in Malaysia, Singapore and other countries with their own employment protections. The Meta litigation highlights how AI tools, even when designed without explicit discriminatory intent, can encode and amplify existing workplace inequalities—particularly those affecting women, workers with disabilities, and those exercising family-related leave rights that are increasingly protected across the region.
The plaintiffs' legal team has made an immediate request: preserving the status quo by keeping the workers employed pending arbitration. This demand underscores the irreversible nature of the alleged harm. Once separations become final, employees lose employer-subsidised health coverage during critical periods such as pregnancy, postpartum recovery and active medical treatment. Time-limited leave entitlements expire, unvested equity is forfeited, and for immigrant workers, separation can trigger visa consequences. These stakes explain why the lawsuit seeks injunctive relief rather than focusing solely on financial damages—the workers want their jobs back, recognising that money cannot compensate for lost leave rights or coverage during vulnerable health periods.
The case also illustrates how modern AI systems in workforce management demand different legal frameworks than traditional hiring and termination practices. Meta's algorithmic selection process operates at scale and speed beyond human capacity for individual case review, yet courts and regulators are still developing appropriate standards for holding such systems accountable. The lawsuit's argument—that Meta's system "by systematically recording such absences as reduced performance, falls more heavily on women than on men"—essentially claims that the company automated discrimination by refusing to adjust its metrics for legally-protected conduct. This framing may influence how courts and regulators worldwide assess AI deployment in employment decisions going forward.
As the case develops, it will test whether human decision-making, when informed by inherently biased algorithmic scoring, can shield a company from liability for those biases. Meta's defence that people, not AI, made final decisions may prove insufficient if those people relied on data generated by systems that systematically disadvantaged protected workers. For technology companies operating globally, including those in Asia, the litigation signals that algorithmic employment decisions face heightened legal scrutiny and that design choices—such as failing to pause or adjust metrics for protected leave—can create legal exposure regardless of stated corporate intentions.
