AI Monitoring at Work: Is Your Computer Watching You?

AI monitoring at work is software that watches what employees do — keystrokes, screen activity, messages, location, sometimes tone of voice — and turns it into scores and alerts a manager can act on. In the United States, most of it is legal, particularly on company-owned equipment.

But “mostly legal” isn’t the same as unlimited. Some states require that you be told. Some uses run into anti-discrimination and disability law. And a productivity score that decides who gets promoted or cut can be wrong about you in ways that are hard to argue with unless you’ve kept your own record.

Here’s what these systems actually track, where the legal lines are, and what you can do about it.

Signs You’re Being Monitored

Most monitoring runs quietly. A few things suggest it’s happening:

• Your employment contract, offer letter, or handbook mentions monitoring, acceptable use, or “electronic resources”
• Software you didn’t install appears on your machine, or your laptop runs slower than the hardware explains
• A manager references something specific you’d expect to be private — how long you were idle, what time you logged off, a message you sent
• You get automated nudges about activity levels, break length, or response time
• Your performance review cites metrics nobody explained to you
• Reviews or scores arrive with numbers attached that don’t match the work you actually did
• You’re asked to install an app on your personal phone for work

That last one deserves its own look — see can employers track your personal phone.

If several of these apply, assume you’re being monitored and read on. The practical question usually isn’t whether it’s happening, but how much of it is lawful and what you can do.

What AI Monitoring Actually Tracks

Traditional workplace monitoring recorded what you did. AI monitoring interprets it — it doesn’t just log that you were idle for 40 minutes, it decides whether that made you unproductive.

What these systems collect falls into a few categories:

Screen and device activity. Keystroke counts, application and website use, periodic screenshots, idle time, and how long you spend in each program. Covered in more depth in can your employer monitor your computer while working from home.

Communications. The content of work email and Slack or Teams messages, and increasingly the metadata around them — who you talk to, how fast you reply, whether you message outside working hours. Some tools run sentiment analysis, scoring messages for tone or “disengagement.”

Location. GPS on company vehicles, work phones, and field apps that log where you are and how long you stay. See can companies track your location.

Biometrics and physical presence. Badge scans, fingerprint and facial recognition at entry points, and in some workplaces wearables that track movement or time away from a station. A handful of states regulate biometric data specifically.

Video and audio. Cameras in the workplace and, on remote teams, webcam access during working hours.

Productivity scores. The output layer. Systems combine several inputs into a single number or ranking and compare you against your coworkers.

That last category is the one that matters most, because the score is what a manager actually sees. You may never be told how it’s calculated, what counts against you, or how you compare. A decision gets made about your work by a system whose reasoning nobody explains — including, often, the manager relying on it.

Why Employers Use It

Understanding the motive helps you judge whether the monitoring in your workplace is proportionate.

• Productivity measurement. The stated reason most of the time, and the shakiest — activity is easy to count and a poor proxy for value.
• Security and compliance. Regulated industries have real obligations around data handling, and monitoring is part of how they meet them.
• Remote work visibility. Managers who can’t see their teams often replace observation with measurement.
• “Objective” performance data. Numbers feel fairer than a manager’s impression, and in some ways they are. But a metric that measures the wrong thing consistently is not more objective than a human — it’s just harder to argue with.

None of these is inherently improper. The question is whether what’s collected matches the reason given, and whether anyone told you.

The Risks to Employees

It measures what’s countable, not what matters. The hour spent calming an angry client doesn’t register. Neither does thinking through a hard problem away from your desk, or helping a coworker who was stuck. Activity metrics reward visible busyness, which is not the same as good work.

It gets people wrong. Systems flag legitimate activity as a problem all the time — the employee who reads carefully before responding looks slow; the one whose work is mostly phone calls looks idle. Left uncorrected, a misreading becomes the record.

It can carry bias. A system trained on past performance data learns the patterns in that data, including any that disadvantaged particular groups. That’s where monitoring stops being a privacy question and becomes a discrimination one.

Constant observation has a cost. Sustained surveillance is associated with higher stress and lower trust. People being scored tend to optimize for the score, which is rarely what the employer actually wanted.

The data itself becomes a liability. Monitoring produces a detailed archive about you — where you were, what you typed, who you spoke to. It sits on a server and is only as secure as the vendor holding it. Employers rarely say how long it’s retained, who can access it, or what happens to it in a breach or an acquisition. Unlike a password, you can’t reset a year of keystroke logs. It’s also discoverable: records collected for “productivity” can be read later in an investigation or a lawsuit for something else entirely. What your employer can gather and keep is covered in what information can your employer collect about you.

What Your Legal Rights Are

On company equipment, employers have wide latitude

In the U.S., an employer generally may monitor activity on devices, accounts, and networks it owns. Work email, company laptops, and employer-issued phones carry very little expectation of privacy. The federal Electronic Communications Privacy Act restricts interception of communications but contains broad exceptions for employers monitoring in the ordinary course of business, and for monitoring you’ve consented to — consent you likely gave by signing the handbook.

Personal devices are a different question, covered in can your employer install monitoring software on your personal computer.

Some states require notice

This is where the rules actually differ. New York requires private employers to give written notice of electronic monitoring to new hires and post it conspicuously. Connecticut and Delaware have their own notice requirements. California grants employees rights to know what personal information is collected about them and why, and biometric data gets specific treatment in states including Illinois, Texas, and Washington.

One-party and two-party consent rules also matter for audio. In two-party consent states, recording a conversation without everyone’s agreement can be unlawful — see can your employer record you at work.

Check your own state. Monitoring law is changing fast, and a rule that didn’t exist three years ago may apply to you now.

If your employer is based outside the U.S.

American workers have weaker monitoring protections than employees in much of the world. If your company is headquartered in Europe, or you work for a multinational with EU operations, the General Data Protection Regulation may shape how your data is handled.

GDPR is stricter in ways that matter: an employer generally must have a legitimate business reason for monitoring, tell employees it’s happening and why, collect no more than the purpose requires, and honor requests from employees to see the data held about them. “We monitor everything because we can” is not a defensible position under that framework.

It doesn’t automatically extend to U.S.-based staff, and whether it reaches you depends on where you work and where your data is processed. But some multinationals apply one standard company-wide rather than maintaining two — worth asking which policy covers you.

When AI Monitoring Crosses a Legal Line

The monitoring itself is usually lawful. What’s done with the results is where employers get into trouble.

Disability discrimination. This is the most common problem. An employee with a disability may work in patterns a productivity algorithm reads as underperformance — more breaks, slower typing, time away from the desk. If a reasonable accommodation would explain the pattern, scoring that employee down and acting on the score can violate the ADA. Employers are also expected to consider accommodations in how performance is measured, not only in how work is done.

Disparate impact. A system applied evenly to everyone can still disadvantage a protected group. If your employer’s productivity metric consistently scores older workers, disabled workers, or employees of one race lower, and the employer can’t show it’s job-related and necessary, that can amount to unlawful discrimination even without intent.

Monitoring protected activity. The National Labor Relations Act protects most employees — union or not — when discussing pay and working conditions with coworkers. Surveillance aimed at catching those conversations, or begun in response to them, is a serious problem.

Retaliation. Monitoring that starts or intensifies right after you complain about harassment, request medical leave, or report a safety issue is a familiar pattern. See is this workplace retaliation.

Off-duty and personal-device surveillance. Tracking that continues after hours, or reaches onto a personal phone beyond what you agreed to, is on much weaker ground than monitoring a company laptop during working hours.

If you think a monitoring system contributed to a demotion, a performance improvement plan, or a termination — and one of these applies — that’s worth a conversation with an employment attorney.

How to Protect Yourself

Read the policy. Your handbook or acceptable-use policy says what your employer claims the right to monitor. Most people sign it without reading it. Knowing what it covers tells you both what to expect and when monitoring has gone past what you agreed to. Request a copy in writing if you can’t find it.

Separate personal from work. The cleanest protection available. Don’t use work devices or accounts for personal business, don’t sign into personal email or banking on a company laptop, and think carefully before installing work apps on a personal phone. Once personal activity is on company equipment, it’s generally fair game.

Assume work communication is visible. Write emails and Slack messages as though a manager will read them, because one may. That isn’t paranoia, it’s the actual legal position on company systems.

Keep your own record of what you accomplish. This is the most useful thing on this list.

AI monitoring measures what’s easy to count and misses most of what makes someone good at their job. So keep a running note of the real work: projects finished, problems solved, feedback received, numbers moved. A few lines a week is enough.

If a productivity score is ever used against you — in a review, on a PIP, in a termination meeting — that record is what lets you respond with specifics instead of arguing about an algorithm you can’t see. It’s also exactly the documentation that matters if you later believe you were treated unlawfully. Your personnel file is worth requesting too.

Raise concerns in writing, and professionally. If monitoring seems excessive or is affecting your work, a short factual email to HR creates a dated record that you raised it. Ask what’s collected, how long it’s kept, and how it factors into performance decisions. Reasonable questions are hard to hold against you, and it’s generally unlawful to retaliate against you for asking.

Don’t try to defeat the software. Mouse jigglers, VPNs to hide activity, and similar workarounds give an employer a clean, legitimate reason to fire you — and they undercut any complaint you might otherwise have had.

Frequently Asked Questions

Related Articles

• Employee Surveillance Laws
• Can Employers Spy on Remote Workers?
• Can Your Employer Require You to Use Monitoring Software?
• Can Your Employer Install Monitoring Software on Your Personal Computer?
• Can Your Boss Read Your Emails?
• Can Employers Monitor Slack or Microsoft Teams Messages?
• Can Companies Track Your Location?
• What Information Can Your Employer Collect About You?
• Internet Privacy at Work

Helpful Outside Resources

• EEOC — Prohibited Employment Policies and Practices
• EEOC — How to File a Charge of Discrimination
• EEOC — Time Limits for Filing a Charge
• U.S. Department of Labor — Workers’ Rights

The Bottom Line

On equipment your employer owns, assume you’re being monitored. That’s the realistic starting position, and in most of the country it’s legal.

What’s worth your attention isn’t the monitoring itself but what gets done with it. Know what your state requires your employer to disclose. Keep your personal life off company devices. And keep your own record of what you actually accomplish — because the day a score is used to justify a decision about your job, that record is the only version of events that’s yours.

If monitoring has contributed to a demotion, a performance plan, or a termination, and you think a protected characteristic or a complaint you made is the real reason, talk to an employment attorney before the filing deadlines run.

Disclaimer

This article is for general informational purposes only and isn’t legal advice. Monitoring and privacy laws vary significantly by state and are changing quickly. If you believe workplace monitoring has been used against you unlawfully, consult a licensed employment attorney in your state for guidance specific to your circumstances. Please read our Terms and Conditions.