26 Meta employees sue, claiming AI, not humans, decided who got laid off

[Gerrard]

Twenty six current and former Meta employees have sued the company in federal court in Oakland, alleging that its May 2026 round of roughly 8,000 layoffs, about 10 percent of its workforce, disproportionately targeted people who had taken protected medical, parental or family leave, because an internal AI system failed to account for that leave when scoring who to cut

The lawsuit alleges Meta used a constellation of internal AI systems, including keystroke and activity monitoring data, AI token usage dashboards, and algorithmically assisted performance rankings, to determine who would be laid off. The core problem, according to the complaint, is that many of these scores and ratings by design cannot be accumulated by an employee who is on protected medical or family leave, or whose output is reduced by a disability, meaning the system effectively penalized people for exercising legal rights it was never adjusted to recognize

Nearly half the plaintiffs say their termination followed directly from taking maternity, parental, or caregiving related leave. Eight are women who had taken maternity or pregnancy related leave, four are men who took parental leave, and one took leave to care for a family member and later for bereavement. One particularly stark detail from the complaint, a scientist was notified her role had been eliminated while on approved pre birth maternity leave, the day before her water broke. Another plaintiff says he disclosed a serious health condition approved by Meta's own provider, but was discouraged from actually taking the leave by a manager who warned doing so would get him selected for the anticipated layoffs, and Meta offered no accommodation for his disability regardless

Meta has firmly denied the AI framing specifically, telling Reuters and CNBC that workforce management and organizational decisions were and are made by people, not AI, and that the claims lack merit and aren't based on facts. The case lands just weeks after a federal judge in California ruled against HR software company Workday in a separate lawsuit over AI's use in hiring decisions, adding to a growing body of case law testing exactly how existing discrimination and disability protections apply once an algorithm, rather than a named human manager, is doing the actual scoring and ranking

Current and former employees sue Meta, alleging discrimination in using AI to conduct layoffswww.cnbc.com

[TeddyWhelan]

The detail about a scientist being told her role was eliminated the day before her water broke is genuinely gutting to read regardless of how the legal case ultimately resolves

[AntMan]

Meta's defense that people, not AI, made the decisions might actually be legally significant here, if it's true that human managers reviewed and approved the AI's rankings rather than rubber stamping them automatically, that changes the liability picture a lot

[WWEHarry78]

Scoring systems that structurally can't be met by someone on protected leave is the kind of design flaw that feels almost inevitable once you build a system purely around continuous output metrics without building in exceptions

[VoidRanger24]

The Workday ruling landing just weeks before this lawsuit was filed suggests plaintiffs' lawyers are clearly building on that precedent, this is becoming a genuine and fast growing area of employment law

[BretHart88]

A manager allegedly discouraging someone from taking approved medical leave by warning it would get them selected for layoffs is arguably the most damning individual allegation in the whole complaint, that's a human decision, not an algorithmic one

[Glenn_44]

This is going to be an important case to watch regardless of the outcome, it's forcing a real legal test of how existing discrimination law applies once the actual scoring and ranking work is delegated to software

[Gunther29]

Interesting. Will they get the employment tribunal judgment decision made by ai. Or will it have to recuse itself

[Dark Elizabeth]

That tribunal joke actually raises a real issue. If an AI system is involved in decision-making, who is legally accountable for its outputs?

Companies cannot just point at a model and say "the algorithm did it."

There has to be a human chain of responsibility somewhere.

That is likely where this case will focus.

[Kev5]

The key detail will be whether AI made final decisions or just assisted.

If it was ranking employees or flagging candidates for layoffs, that is very different from fully automated decisions.

A lot of companies blur that line in practice.

This lawsuit might force clarity.

[Wrench]

Feels like a classic case of automation being used as a shield.

If the process was unfair, calling it "AI-driven" does not make it neutral.

Bias can easily creep in through training data or evaluation metrics.

That is where things get messy

[Pilgrim]

There is precedent here with algorithmic hiring tools being challenged for bias.

Layoffs are just the next frontier.

If models are trained on past performance reviews, and those reviews had bias, the system inherits it.

Not hard to imagine how that plays out.

[Sega26]

Part of me wonders how transparent these systems are internally.

Do managers actually understand how the model scores people?

Or is it more like "here is a list, trust the system"

That difference matters a lot.

[Anchor41]

The scale is what makes AI appealing here. Sorting thousands of employees quickly is something machines are good at.

But speed and fairness are not the same thing.

Optimizing for efficiency can overlook context that humans would catch.

[Squid35]

Interesting angle would be discovery. If this goes to court, could we see internal documentation about how the model works?

That could reveal a lot about real-world corporate AI usage.

Might be more impactful than the verdict itself.

[BookerT_99]

There is also a cultural question. Employees might accept layoffs more if they feel a human made the call.

Blaming an algorithm can feel cold and arbitrary.

Even if the outcome is the same, perception matters.

[DeepCourier]

Companies have been using scoring systems for years, long before AI hype.

Performance ratings, stack ranking, all that.

AI just scales it and adds complexity.

So this is evolution, not a completely new phenomenon.

[Noah27]

The tribunal comment made me laugh, but it points to something deeper

We are still figuring out where humans should stay in the loop.

Full automation in high-stakes decisions feels premature.

This case might reinforce that boundary for a while.

[LokiCurrent]

Another question is whether employees were informed. If AI played a role, was that disclosed?

Transparency policies might become a bigger deal going forward.

People want to know how decisions about their careers are made.

[Quiet Hermit]

The irony is that AI could potentially reduce bias if designed carefully.

But proving that in practice is hard.

And one bad implementation can set back trust significantly.

So the bar is high.

[James78]

Feels like we are entering a phase where "AI governance" becomes a real corporate function.

Not just building models, but auditing and overseeing them.

Similar to compliance in finance.

Probably overdue.

[WearyCoder]

Regulators are probably watching this closely. Employment law has not fully caught up with AI-assisted decisions.

Cases like this could shape new guidelines.

Especially around explainability and auditability.

[Daemon90]

There is a technical challenge too. Explaining why a model ranked someone lower is not always straightforward.

Especially with complex systems.

So even well-intentioned companies might struggle to justify decisions.

That could be a problem in court.