An AI system may have decided who Meta laid off. And it allegedly went after the people on medical leave.

26 of them are now suing.

They filed in federal court this week, saying Meta used internal AI systems, including the token dashboards and keystroke monitoring I wrote about recently, to help pick who got cut when it laid off 8,000 people.

And they say it flagged the wrong people. Workers out on medical, disability, or parental leave.

These are allegations. Nothing's proven, and Meta denies all of it.

But think about how a system like that works.

It scores you on activity. So what does it see when someone's out for chemo? On maternity leave? Home with a dying parent?

A low number. That's it.

The machine can't tell "not working" from "out for a reason the law protects." It sees the dip and marks you down.

The lawyers argue this hit women hardest. Makes sense. Women take most of the leave.

Here's the part I can't get past.

Meta's defense is that "decisions were made by people, not AI."

This is the company that spent a year making AI usage a performance metric. Built the leaderboard. Logged keystrokes to feed the systems.

You don't get to run the company on the number, then say it never mattered the second it lands in court.

Either the metric decided, or the person did. Pick one.

This is what I keep saying about AI and jobs.

The danger was never that the machine wakes up and turns on us. It's that it becomes the excuse.

"The system flagged them" is a convenient place for accountability to disappear.

An algorithm can sort a spreadsheet. It can't answer for what that does to someone's life. Only a person can.

Meta says the claims lack merit, and a court will sort it out.

But if you're putting AI near decisions about people, read this filing.

A number that dings someone for having a baby was never the efficiency win somebody promised in the deck.