A machine doesn't care if you're recovering from surgery or holding a newborn. It only cares about the numbers. If your keyboard is quiet, you're a liability.
That is the terrifying reality central to a groundbreaking federal lawsuit filed in California. A group of 26 anonymous employees recently sued Meta Platforms, claiming the social media giant used biased, automated software to choose who to fire during its massive job cuts. If you liked this article, you might want to look at: this related article.
The core of the dispute? The company supposedly relied on a suite of surveillance tools and artificial intelligence metrics that automatically flagged employees on protected family, pregnancy, or medical leave.
When you're away on approved FMLA leave, your digital output drops to zero. You aren't typing. You aren't querying database engines. You aren't using internal tools. To a manager, that makes sense. You're on leave. But to an algorithm tracking daily performance metrics, it looks like you suddenly became the most unproductive employee on the payroll. For another look on this event, see the recent coverage from Wired.
This is the first major legal challenge of its kind. Meta is accused of using AI to pick employees with medical conditions for layoffs, and the fallout will reshape how corporate America operates.
Inside the black box of corporate surveillance
How does a company like Meta measure your worth? It is no longer about whether your manager likes your work or if you hit your project deadlines. According to the 71-page lawsuit, Meta built a complex network of automated tracking systems to score its staff.
The tools used to judge employees paint a dark picture of modern office life:
- Metamate: Meta's internal large language model assistant. The system monitored how frequently employees interacted with it.
- The Second Brain: An AI system that tracked worker communications, messages, and shared documents to measure active collaboration.
- Token-usage dashboards: Dashboards that tracked how much computing power or API tokens an engineer consumed while writing and testing code.
- Keystroke and activity monitoring: Background software that recorded keyboard strokes, mouse movements, browser history, and email traffic.
If you are a software engineer working late nights, these tools record a steady stream of activity. But what happens if you break your wrist? Or what if you're a scientist put on bed rest due to a high-risk pregnancy?
Your mouse stops moving. Your token usage hits zero.
The lawsuit alleges that Meta's layoff algorithms didn't pause to ask why someone's productivity dropped. The code treated a pregnant engineer on maternity leave the same way it treated a low-performing worker who was slacking off. The results were brutal. One plaintiff, a scientist on approved pregnancy leave, was notified of her layoff just two days before she gave birth. Another engineer saw his performance rating slashed simply because he took approved time off to recover from an injury.
The defense of human decision making
Meta has pushed back hard against these claims. Company spokespeople argue that the lawsuit lacks merit, stating that "workforce management and organizational decisions were and are made by people, not AI."
It is a clever defense, but it highlights a massive grey area in employment law.
Even if a human manager signs the final termination paperwork, who actually built the list? If a manager is handed an algorithmically generated spreadsheet of the bottom 10% of performers and told to cut them, did a human make that decision? Or did they just rubber-stamp the machine's bias?
Under the hood, managers are increasingly pressured to trust data over their own intuition. In a hyper-optimized tech company, going against an algorithm's recommendation is risky for a middle manager. If they save an employee whom the AI flagged as low-performing, and that employee misses a deadline, the manager takes the fall.
The pressure to trust the machine is intense.
The legal reality of algorithmic bias
This lawsuit is a direct challenge to the idea that companies can hide discrimination behind a curtain of proprietary code.
Under federal laws like the Americans with Disabilities Act (ADA) and the Family and Medical Leave Act (FMLA), employers must provide reasonable accommodations and protect workers who take leave. If an employer uses an algorithm that inherently penalizes people for using those rights, that algorithm is illegal.
The plaintiffs also point to newer state and city laws. For example, local regulations in New York City and California require companies to audit their automated employment decision tools for bias before using them. The lawsuit claims Meta failed to perform these critical bias tests, letting an unchecked algorithm run rampant on its staff database.
The immediate goal for these 26 workers is to get a preliminary court ruling to pause their layoffs, which were scheduled to finalize on July 22, 2026, while they fight their cases in private arbitration.
Why this matters to every worker
This isn't just a Silicon Valley problem. Surveillance software is cheap, easy to install, and spreading rapidly to other industries. If you work in finance, customer service, or logistics, your employer is likely tracking similar metrics.
If this case fails, it sets a dangerous precedent. It tells companies that they can bypass decades of civil rights protections simply by outsourcing their firing decisions to a third-party algorithm. It gives them plausible deniability.
So, what can you do to protect yourself?
Keep your own receipts
Never rely on your employer's systems to prove your worth. If you take approved medical leave, keep physical or personal digital copies of your approval letters, doctor's notes, and emails with HR.
Document your performance early
Keep a running log of your contributions, positive feedback, and completed projects. If you are suddenly hit with a low productivity score after returning from leave, you need hard evidence to show your actual value.
Understand your rights
FMLA and ADA protections are clear. If you feel you are being targeted or managed out after requesting an accommodation, consult an employment lawyer immediately. Do not wait for the layoff notice to arrive.
The tech industry wants us to believe that algorithms are objective and fair. This lawsuit proves they are often just a digital cover for old-fashioned discrimination. If we let machines decide who is worthy of a job, we lose the human empathy that keeps workplaces functional. Let's make sure the law catches up to the code before it is too late.
