MENLO PARK, California, July 19 — Meta is facing a legal challenge over its use of automated AI systems to rank employees for termination, a case that pulls back the curtain on how the company operationalised its sweeping workforce reductions. The lawsuit, filed against the Menlo Park-based social media giant, alleges that the company’s algorithm-driven layoff process failed to account for workers on legally protected leave — a failure that, if proven, cuts to the heart of employment law protections that have long resisted automation.
Meta has conducted multiple layoffs since 2022, including a wave in May that affected thousands of employees. The company has publicly framed these reductions as part of a strategic restructuring, the kind of corporate language that tends to land in press releases and all-hands memos.
Under the hood, however, the technical reality was something more mechanical: automated tools that scored and ranked workers, then fed those rankings into termination decisions. The detail that matters here is not just that machines were making cuts, but that the criteria baked into those machines may have been blind to legally mandated protections.
The case is one of the first to specifically challenge the failure of AI systems to account for workers on protected leave — a category that traditionally covers medical leave, family leave, military service, and disability accommodations, among others. If the algorithm ranked such workers lower or flagged them for termination without factoring in the protected status of their absence, the legal argument runs, the company effectively outsourced discrimination to software.
As ever in these cases, the opacity of the system itself becomes part of the problem: if the company does not know exactly how its AI ranked workers, it cannot defend those rankings in court. The use of algorithms in layoffs is a growing trend among large tech firms, a development that observers have watched with increasing unease. The logic is straightforward enough: when a company needs to cut thousands of roles, manual review by human managers is slow, inconsistent, and expensive.
Automated ranking offers speed, scale, and a veneer of mathematical objectivity. The technical reality, however, is that these systems are only as fair as the data and criteria they are fed — and that data is often historical, reflecting existing biases, managerial preferences, and gaps in the record that no algorithm can self-correct. The wry observation that suggests itself is that Silicon Valley, having spent years building AI to automate every conceivable business function, is now discovering that the hardest thing to automate may be fairness.
This lawsuit puts that proposition to a direct legal test. If the court finds that Meta’s algorithm unlawfully disadvantaged workers on protected leave, it will set a precedent that ripples through every tech company currently building or buying similar tools.
That is a lot of companies. The trend is not limited to Meta; the broader industry has been quietly adopting algorithmic HR tools for hiring, performance review, and now termination, often with little public scrutiny.
The technical question at the heart of the case
The lawsuit does not merely allege that Meta used AI in layoffs — that much the company has never denied. The core allegation is that the AI system was designed or deployed in a way that systematically failed to account for protected leave status. This is a technical failure as much as a legal one.
An algorithm that ranks workers by productivity metrics, for example, will naturally penalise anyone whose absence from work was legally protected. The algorithm does not know it is discriminating; it only knows that absenteeism correlates with lower output.
The detail that matters is whether Meta built in any mechanism to flag or override such scores for workers on protected leave. The lawsuit says it did not. The timing is significant.
Meta’s May layoff wave was one of the largest in its history, part of a cycle of workforce reductions that began in 2022 and has continued through 2026. Each round has been justified as part of a strategic restructuring — the company’s public framing — but the cumulative effect has been a dramatic reduction in headcount.
The lawsuit now forces a reckoning with the internal mechanics of those cuts, and the question of whether the company’s strategic restructuring was also a legally actionable one.
What comes next
The case is in its early stages, and the court has not yet ruled on the central allegations. What makes it significant beyond the immediate parties is its potential to establish judicial standards for the use of AI in employment decisions. US employment law has long held that employers are responsible for discriminatory outcomes regardless of intent — the principle of disparate impact.
If that principle applies to algorithms, then companies cannot hide behind the complexity of their systems. The technical reality of how the AI ranked workers becomes evidence, not excuse.
For now, Meta has declined to comment on the specifics of the litigation, maintaining its public posture that the layoffs were a necessary part of the company’s restructuring. The lawsuit, however, has already accomplished something that the company’s press releases never did: it has opened the black box, if only a crack, and revealed the automation beneath. The detail that matters, as ever, is not the technology itself but what it was allowed to do, and who it was allowed to do it to.
That is a question for the courts, and for an industry that has yet to decide whether fairness is a feature or a bug.
