AI Employee Monitoring: Is Your Company Watching — and Ranking — You?
Most of the conversation about AI at work is about AI as a tool — the thing you use to draft the email or crunch the numbers. But there’s a quieter version of the story that just landed in the headlines: AI as a lens your employer points back at you.
In July 2026, a group of 26 current and former Meta employees sued the company, alleging that its mass layoffs — roughly 8,000 people as Meta remade itself into an “AI-first” company — were driven by AI systems that scored, ranked, and selected workers for termination. The complaint alleges the company leaned on internal AI tools, keystroke and activity-monitoring data, “AI token-usage dashboards,” and algorithmically assisted performance rankings — and that because those metrics “by design, cannot be accumulated by an employee who is on protected medical or family leave,” the cuts fell disproportionately on people who were out on leave. (These are allegations; the case hasn’t been decided.)
Whatever a court eventually finds, the lawsuit is a useful window into something that’s already widespread and rarely explained in plain terms.
What “algorithmic management” actually means
The umbrella term for this is When software — not just a human manager — measures, evaluates, and makes or recommends decisions about workers. Instead of your boss forming an impression of your work, a system collects data about your activity, turns it into scores or rankings, and feeds those into real decisions: who gets a good review, who gets a shift, who gets promoted, who gets cut. Gig platforms pioneered it; it’s now spreading into ordinary office jobs. . It’s not science fiction and it’s not rare. By industry surveys, a large majority of U.S. employers now run some form of digital monitoring, and a majority use AI-driven analytics to measure productivity or behavior — a shift from simply watching to automatically judging.
What companies can actually see
The raw material for all of this is The routine collection of data about what you do on work devices — which apps and websites you use and for how long, when you’re active, keystroke and mouse activity, sometimes periodic screenshots, and increasingly how much you use approved AI tools. Much of it runs quietly in the background, often disclosed only in a policy you clicked “agree” on when you were hired. . Individually, none of these signals mean much. The change AI brings is stitching them together into a single number — a productivity score, a performance rank, a percentile — that looks objective and travels into consequential decisions.
That “looks objective” part is exactly where it gets dangerous.
Why the numbers can lie
An activity score measures activity, not value — and those aren’t the same thing. The Meta suit points straight at the gap: metrics like keystroke volume and tool-usage dashboards can’t be earned by someone on medical or parental leave, or by someone whose disability changes how they work, or by the person whose most valuable contribution that quarter was an hour of judgment in a meeting that no dashboard recorded.
This is the old “garbage in, garbage out” problem wearing a lab coat. When a system ranks people on what’s easy to measure, it quietly punishes everyone whose real worth is hard to measure — and it does it with a confidence and a veneer of neutrality that a human manager’s gut never had. A biased hunch can be argued with. A number that “the algorithm produced” often isn’t.
What you can reasonably do
You can’t opt out of your employer’s systems, but you’re not powerless either:
- Know what’s actually collected. Your monitoring policy is usually in the employee handbook or the consent you signed at hire. Read it. Knowing whether screenshots, keystrokes, or AI-usage are tracked is the difference between guessing and knowing.
- Don’t let a dashboard be the only record of your value. Keep your own — a running “brag document” of what you shipped, decided, unblocked, and mentored. When your worth is contested, the version the algorithm captured shouldn’t be the only one in the room.
- Understand that “looking busy” metrics are gameable and hollow. The goal isn’t to farm keystrokes; it’s to make sure your genuinely valuable work is visible to humans, not just legible to software.
- Know your protections. Medical and family leave, disability accommodations, and — depending on where you live — laws about monitoring and automated decisions are real. This is contested legal ground (the Meta case is one of several testing it), so if something feels wrong, it’s worth asking about rather than assuming it’s settled.
The takeaway
The headline lesson of the Meta lawsuit isn’t “AI is evil.” It’s that AI at work runs in two directions: the tool you point at your tasks, and the lens your employer points at you. The second one is already here, mostly undiscussed, and built on the shaky assumption that what’s easy to measure is the same as what matters. You don’t need to panic about it. You need to be aware of it — know what’s tracked, keep your own record, and remember that a confident-looking number is still just a number. Being legible to software is not the same as being valuable, and the gap between them is where your judgment, and your rights, still live.
Sources: The lawsuit details are from CBS News and Fortune; all claims about Meta are allegations from a complaint filed in the U.S. District Court for the Northern District of California and have not been proven. Figures on the prevalence of workplace monitoring and AI-driven productivity analytics are from industry surveys compiled by HRStacks.
Related: AI Anxiety at Work: You’re Stressed and More Productive. Both Are True. on the job-security worry underneath all this, and Write a One-Page AI Policy Your Team Will Actually Follow if you’re the one setting the rules.