Meta’s AI Layoff Lawsuit: When Algorithms Punish Sick Employees for Taking Leave
Twenty-six former Meta employees have filed a lawsuit alleging that AI-driven performance metrics used during the company's 2026 layoffs failed to account for medically protected absences, effectively penalising workers for exercising their legal rights. The case could set a significant precedent for how algorithmic HR tools are regulated across the industry.
When a corporation lays off thousands of workers, the question that usually follows is: who decided, and how? At Meta, that question has now landed in a courtroom — and the answer, according to twenty-six former employees, is deeply troubling. They allege that an AI-driven performance scoring system quietly penalised workers who were on medically protected leave, effectively letting an algorithm make consequential human decisions without any adjustment for legally protected absences.
This story, first reported in the Los Angeles Times and highlighted by Mindstream, sits at the intersection of two of the most contentious conversations in technology today: the rapid adoption of AI in corporate decision-making, and the rights of workers who stand in its path.
What the Lawsuit Actually Claims
The lawsuit filed by the twenty-six workers is precise in its language. As Mindstream reports, the court filing directly states: “The result was that employees who took protected leaves were disproportionately selected for layoff, based on scoring that not only failed to account for their protected leaves, but in effect penalized the employees for exercising their legal rights to these leaves.”
Breaking that down, here is what the workers are alleging:
- Meta used AI-informed performance metrics during its 2026 round of layoffs.
- Those metrics tracked productivity in ways that did not exclude periods of approved medical leave, pregnancy leave, bereavement leave, or disability-related absences.
- Employees who had exercised legally protected rights to take time off ended up with lower performance scores as a direct consequence.
- Those lower scores then fed into decisions about who would be let go.
Under both federal and state law in the United States, medical and family leave is a protected right. If a scoring system structurally disadvantages workers for exercising that right — even without any explicit discriminatory intent — it can constitute illegal employment practice. That is the crux of the legal argument, and it is why legal observers are watching this case closely.
Meta has pushed back firmly. The company’s stated position, as reported by Mindstream, is unambiguous: “Workforce management and organizational decisions were and are made by people, not AI.” That defence is itself revealing. By distancing the company from any suggestion that an algorithm made final calls, Meta is essentially arguing that human managers retained full accountability. Whether the evidence supports that claim will be central to how this case unfolds.
The Broader Context: Meta’s 2026 Layoffs
This lawsuit does not exist in a vacuum. It is, as Mindstream frames it, a continuation of a story that began earlier in 2026, when Meta laid off around 8,000 employees — approximately 10% of its total workforce. The stated rationale was strategic reallocation: cutting headcount in certain areas to redirect resources toward AI development initiatives.
That kind of pivot is not unique to Meta. Across the technology sector, companies have been using cost-reduction exercises to fund AI infrastructure, framing layoffs as a necessary precondition for building the future. The logic is internally consistent, even if it is cold comfort for those who lose their jobs.
What made Meta’s approach particularly contentious, according to Mindstream, was the granularity of the surveillance involved. The company reportedly used AI to monitor employee activity down to daily keystrokes. Separately, leaders encouraged employees to train personal AI agents — described as “second brains” — that could effectively replicate their job functions. Reading between the lines, this created a situation where workers were simultaneously being evaluated by algorithmic metrics and being asked to build the tools that might replace them.
For employees on medical leave during any part of this period, that dynamic was especially punishing. Their absence from active work — an absence they were legally entitled to — registered as reduced output in a system that apparently did not know, or did not care, to account for it.
Why This Case Could Set a Precedent
The legal stakes here extend well beyond Meta and its former employees. If the courts find that AI-assisted performance systems can constitute discriminatory practice when they fail to adjust for protected absences, it would send a clear signal to every large employer using similar tools.
The use of algorithmic performance management is widespread. Warehouse workers tracked by fulfilment quotas, customer service representatives scored by call metrics, software engineers evaluated by code commit rates — across industries, productivity is increasingly measured by automated systems. Most of those systems were designed to track output, not to understand the human context behind fluctuations in that output.
A ruling against Meta would not necessarily prohibit algorithmic performance management. But it would create a legal obligation to audit those systems for bias against protected classes — and that audit requirement alone could reshape how companies deploy these tools.
From an Indian workplace perspective, this matters too. As multinational technology companies with large engineering and operations workforces in India increasingly adopt global HR technology platforms, the standards set by US litigation tend to travel. Indian employees working for these firms — or for domestic companies adopting similar tools — have a stake in how these legal questions resolve.
The Fundamental Tension AI Cannot Resolve on Its Own
At its core, this case is about a question that no algorithm can answer by itself: what does it mean to treat workers fairly?
Performance metrics are seductive because they appear objective. Numbers feel neutral. But every metric encodes a set of assumptions about what counts as valuable work and what does not. When those assumptions go unexamined — when a system is built to measure keystrokes and commit counts without asking whether the person being measured was recovering from surgery or managing a pregnancy — the resulting scores are not neutral. They are quietly discriminatory.
Meta’s defence rests on the claim that humans made the final decisions. But if those humans were working from AI-generated scores that were structurally biased against workers on protected leave, the human sign-off becomes, at best, a formality. The algorithm shaped the decision set; the manager merely chose from options the system had pre-filtered.
This is precisely the kind of accountability gap that critics of algorithmic management have warned about for years. When something goes wrong, the company can point to human decision-makers. The human decision-makers can point to the data. And the data, being an abstraction, cannot be held responsible.
What Comes Next
The lawsuit is ongoing, and Meta has denied any illegal or unethical conduct. The outcome will depend significantly on what discovery reveals about how the performance scoring system was actually built, what data it used, and whether anyone at Meta raised concerns about its treatment of protected absences before the layoffs took place.
For workers — in the United States or elsewhere — the case is a reminder that understanding the systems being used to evaluate you is no longer just a matter of career strategy. It is becoming a matter of legal rights. And for companies deploying AI in HR functions, it is a warning that optimising for efficiency without auditing for fairness carries real legal and reputational risk.
As Mindstream puts it with pointed simplicity: Meta may call it optimisation. But workers are calling their lawyers.
