Your Boss Is Watching. And the Machine Is Taking Notes.
- Michael Routhier
- 4 days ago
- 7 min read
I want to start with something they told us.
When AI started entering the workplace a few years ago, the story was simple and it was everywhere. AI is going to free workers from drudgery. AI is going to eliminate the boring tasks so human beings can focus on the creative ones. AI is going to make your job easier, your days less stressful, and your work more meaningful.
That was the promise.
Here is what's actually happening.
Employers are using AI to monitor when you arrive at your desk, not by the minute, but by the second. They are using it to count your keystrokes, track your mouse movement, analyze your email tone for signs of disengagement, flag your Slack messages for keywords that suggest you're thinking about leaving, score your calls against an algorithm, and generate a productivity rating that determines your future at the company, without a single human manager ever reviewing what the number actually means.
The machine is not helping you do your job better.
The machine is being used to build a case against you.
What the Research Actually Shows
The numbers on workplace AI surveillance are not coming from advocacy groups or union activists. They are coming from the companies themselves, and from the academic researchers who study them.
A 2026 study found that the use of AI-powered employee monitoring tools has increased by over 65 percent since 2023. Not pilot programs. Deployed, active systems, running on your work computer right now if you work for a mid-size or large company, whether you know it or not.
What are these tools doing?
The list is longer than most workers realize. Continuous screen capture; some systems take screenshots every 30 seconds and run them through AI to analyse what you're looking at. Keystroke and mouse tracking to generate "productivity scores". Email and message sentiment analysis; AI reading the emotional tone of your communications and flagging anything that suggests low morale, conflict, or flight risk. Video monitoring that uses AI to measure how often your face appears on camera during video calls. Location tracking through building access systems that logs when you enter rooms, how long you stay, and who you're near. Some systems even measure the "energy" of your writing; flagging if your messages are getting shorter, more terse, or less collaborative over time.
None of this data is being used to help you. It is being used to evaluate you, rank you, and in many cases, to build an automated justification for managing you out of the organization, without a human manager ever having to sit down and tell you why.
The Dictator Comparison Nobody Wants to Make
I made it in my post about dictatorships using AI to control their citizens. I'll make it again here, because the parallel is too important to ignore.
I wrote about how authoritarian governments use AI surveillance to monitor dissent; to flag citizens who express the wrong opinions, associate with the wrong people, or simply fail to display sufficient enthusiasm for the regime. I wrote about China's Social Credit System, about Iran's AI cameras, about the way surveillance infrastructure exported from authoritarian states becomes a template for every regime that follows.
Here is the question I didn't ask loudly enough then, but I'm asking now;
What is the functional difference between a government that uses AI to monitor its citizens for signs of disloyalty and an employer that uses AI to monitor its workers for signs of disengagement?
The architecture is identical. The justification is different. The outcome, a person being scored, flagged, and acted against based on algorithmic surveillance they didn't consent to and can't appeal, is the same.
The researchers studying authoritarian AI surveillance documented something they call the autocrat's calibration dilemma, the impossibility of setting a surveillance threshold that doesn't either sweep up innocent people or miss genuine threats. The same dilemma exists in workplace AI monitoring. Set the productivity score threshold too low and you flag your best employees for having a bad week. Set it too high and the disengaged workers who have learned to game the metrics sail through untouched.
The surveillance doesn't work the way it claims to. But by the time that becomes clear, real people, real workers with families and mortgages and careers, have already been quietly managed out by an algorithm that nobody was required to explain or defend.
The Lie That Was Told With a Straight Face
There is a specific kind of dishonesty that makes me angrier than outright deception. It's the kind where the people doing the harm genuinely believe their own framing.
The employers deploying these systems are not, for the most part, twirling their moustaches in a boardroom somewhere. They genuinely believe they are optimizing their organizations. They genuinely believe that more data leads to better decisions. They genuinely believe that if you have nothing to hide, you have nothing to fear from the keystroke counter.
And that belief, that sincere, comfortable, never-examined belief, is more dangerous than cynical dishonesty. Because it means there's no one to negotiate with. There's no bad actor to identify and remove. There's just a system that has been declared neutral, objective, and efficient, and therefore beyond question.
Marcus Aurelius wrote; "What is it fundamentally that this thing is for, not what it claims to be for, but what it is actually for?"
The productivity monitoring system claims to be for helping managers understand team performance. What it is actually for is shifting power permanently and completely from workers to employers; replacing the messy, human, negotiated relationship between a manager and an employee with an automated verdict that requires no conversation, no accountability, and no appeal.
That is not optimization. That is control. And the word for a workplace where your every movement is monitored, scored, and used to determine your future, without transparency, without recourse, and without your meaningful consent, is not an efficient workplace.
It is a surveillance state with a dress code.
And What About You?
I talk to adults over 55 every week in this community. Many of you are still working, and the ones who are have told me, again and again, that the workplace has changed in ways that feel deeply wrong but are hard to articulate.
Now you can articulate it.
The reason your performance review no longer feels like a conversation is that it isn't one. The reason your manager seems to know things about your work patterns that you never told them is that an algorithm told them first. The reason you feel watched, always slightly watched, even when nobody is in the room, is that you are.
And there is a specific cruelty here for workers over 55 that I want to name directly.
AI productivity monitoring systems are trained on data. The data reflects what the system's designers think productivity looks like. And what they think productivity looks like is, in the vast majority of cases, a 32-year-old in a fast-paced environment producing high-volume output with fast response times.
Older workers, who bring institutional knowledge, mentorship, relationship skills, judgment, and the kind of deep expertise that doesn't show up in a keystroke count, are systematically disadvantaged by these metrics. Not because they are less productive. Because the machine doesn't know how to count what they're actually contributing.
And when the algorithm flags them, the employer has a number to point to. Not a conversation. Not an evaluation. A number. Clean, objective-looking, and completely divorced from reality.
That is not a fair system. That is a system designed, whether intentionally or not, to push older workers out.
What You Can Do
Here is what I want you to actually do with this information.
One: Find out if your employer is monitoring you. In Canada and many US states, employers are required to disclose monitoring in writing. Ask your HR department directly; what data is collected about my work activity, how is it stored, who has access to it, and how is it used in performance evaluation? If they can't answer that clearly, you have your answer.
Two: Document everything yourself. If you are in a performance improvement process, or if your role is being restructured, or if you have reason to believe AI monitoring is being used against you, start keeping your own records. Emails. Meeting notes. Outputs. Timestamps. A human paper trail to counter an algorithmic one.
Three: Talk to a union or employment lawyer. In most jurisdictions, the legal framework around AI-based employment decisions is genuinely unsettled. Courts are still working through what employers are and are not allowed to do. That unsettled landscape is actually an opportunity; if you've been harmed by an AI-driven employment decision, there may be legal recourse that didn't exist two years ago.
Four: Make noise. The most important thing that needs to happen in AI workplace governance is public awareness. These systems operate in silence because the workers subject to them don't know they exist. When you know, tell people. Tell your colleagues. Tell your family. Tell your representatives.
The dictator's surveillance state depends on silence. So does the employer's.
Epictetus, a man who was literally owned by another human being, who understood what it meant to have every movement controlled and recorded, said; "Freedom is the only worthy goal in life. It is won by disregarding things that lie beyond our control."
But here is the thing about AI monitoring; it is not beyond our control. Not yet. The legal frameworks are still being written. The public awareness is still forming. The companies deploying these systems are still sensitive to scrutiny because the scrutiny is still new.
This is the moment. Not after the systems are normalized. Not after every performance review runs through an algorithm and everyone has forgotten what it felt like to be evaluated by a human being who knew your name.
Now.
The machine is taking notes.
It's time to take some of your own.
Michael Routhier is the founder of Tech 4 Grown-Ups, a platform dedicated to honest, unfiltered digital literacy for adults 55 and over. The Virtuous Machine is a series exploring the ethics, power, and human cost of artificial intelligence. Find everything at tech4grownups.com.
Comments