Can Your Employer Use AI to Monitor, Evaluate, or Discipline You?
Learn how employers may use AI to monitor productivity, evaluate performance, influence employment decisions, and what federal, state, and local workplace protections may apply.
In this article
- What Employers Can Use AI to Do
- AI Monitoring Does Not Automatically Mean Your Employer Is Breaking the Law
- Can AI Be Used to Evaluate Your Performance?
- AI Can Influence Hiring, Promotion, Pay, Discipline, and Termination
- Federal Employment Protections Still Apply When AI Is Involved
- Disability Can Be an Important Issue
- Workplace AI Can Also Raise Labor-Law Issues
- There Is No Single Federal "AI at Work" Rule
- New York City Has Specific Rules for Automated Employment Decision Tools
- Illinois Has Specific AI Employment Protections
- Colorado Has an AI Framework for Consequential Decisions
- California Is Adding Further Worker Protections, But Timing Matters
- What to Do If You Think an AI Decision Was Wrong
- 1. Write down the employment decision
- 2. Preserve relevant communications
- 3. Look for information about automated systems
- 4. Ask focused questions
- 5. Compare the decision with your actual work
- Pay Particular Attention to Discrimination or Retaliation
- Keep an AI Employment Decision in Perspective
- What Employees Should Watch for Going Forward
- A Practical Way to Think About AI Workplace Rights
- The Bottom Line for U.S. Employees
- Conclusion
Artificial intelligence is increasingly being used in workplaces for tasks that used to be handled entirely by people. An employer may use software to analyze productivity, screen job applications, evaluate performance, monitor workplace activity, recommend promotions or pay decisions, or help determine who should be disciplined or terminated.
But the fact that an employer uses AI does not automatically make the practice illegal, and there is no single federal law that gives every U.S. employee the same set of rights whenever AI is used at work.
The important question is how the AI is being used, what decision it influences, what information it relies on, and which laws apply to the employee and employer.
For workers, that distinction matters. An AI system might simply organize information for a manager, recommend an action, or play a much larger role in the final decision. Federal employment protections can apply when AI produces discriminatory results, while some states and cities have added specific rules concerning automated employment decisions, notice, human review, or employee rights.
Here's how to understand the issue and what you can do if you believe AI has affected your job.
What Employers Can Use AI to Do
Workplace AI is broader than a chatbot or an application that generates text. Employers can use automated systems to collect, analyze, rank, predict, or recommend information about applicants and employees.
The Equal Employment Opportunity Commission (EEOC) identifies several ways workers may encounter AI and automated technology. These include resume screening, recorded interview evaluation, workplace surveillance, productivity monitoring, location tracking, analysis of facial expressions or voice, email and keystroke monitoring, customer-survey analysis, promotion and pay recommendations, and decisions involving layoffs or termination. EEOC
Depending on the employer and the job, AI may therefore be involved in several different stages of the employment relationship.
The important point is that these uses are not legally identical. A system that helps a manager organize information is different from a system whose output materially influences a consequential employment decision.
That is why simply asking, "Is my employer using AI?" may not be enough.
A more useful question is: What is the AI doing, what information is it using, and how much influence does it have over the employment decision?
AI Monitoring Does Not Automatically Mean Your Employer Is Breaking the Law
Employers can have legitimate reasons for monitoring workplace activity. Depending on the job and circumstances, an employer may want to measure productivity, protect company systems, manage staffing, investigate misconduct, or understand how work is being performed.
AI can make that monitoring more extensive because software can analyze large amounts of information much faster than a person could.
For example, a workplace system might analyze:
- How long an employee spends on particular tasks
- Computer or application activity
- Email activity
- Keystrokes or cursor activity
- Location information
- Work patterns
- Customer feedback
- Voice or facial characteristics
- Other information generated during work
The EEOC specifically identifies several of these activities as examples of workplace surveillance involving AI or automated technology. EEOC
That does not mean every form of monitoring is prohibited.
Instead, different legal questions may arise depending on what is being monitored and how the information is used. For example, an AI monitoring system may become an employment-discrimination issue if it produces discriminatory outcomes. It may also raise labor-law concerns if monitoring interferes with legally protected employee activity.
Company policies, employment agreements, applicable state laws, local ordinances, and the circumstances of the monitoring can also matter.
For that reason, an employee should be careful about assuming either extreme:
"My employer uses AI, so the monitoring must be illegal."
or
"My employer is allowed to monitor me, so anything the AI does is automatically legal."
Neither assumption is reliable.
Can AI Be Used to Evaluate Your Performance?
Yes. An employer may use AI or automated systems to evaluate aspects of employee performance.
A system could potentially analyze productivity measurements, task completion, customer feedback, work activity, attendance-related information, or other data selected by the employer.
The problem is that a measurement can look objective without necessarily providing a complete picture of someone's actual performance.
Consider a hypothetical employee whose productivity score falls because an automated system counts time spent away from a computer. That score might not account for meetings, training, conversations with customers, equipment problems, work performed away from the computer, or other responsibilities.
An AI-generated score may therefore be only one piece of information rather than a complete assessment of an employee.
This becomes particularly important when the output is used for promotion, pay, discipline, or termination.
Employees should understand the difference between:
AI as an information tool:
The system analyzes information and gives a manager another source of information to consider.
AI as a recommendation tool:
The system recommends an action, such as flagging an employee for additional review.
AI as a decision-making system:
The system's output has a much more direct role in determining the employment outcome.
The legal consequences can differ depending on the circumstances and applicable law.
There is no universal federal requirement that a human review every AI-assisted employment decision. Some state or local laws, however, can impose additional requirements for particular automated decisions.
AI Can Influence Hiring, Promotion, Pay, Discipline, and Termination
AI does not have to make the final decision itself to matter.
Suppose a manager makes the final decision to deny an employee a promotion, but the manager relies heavily on an AI-generated performance ranking. The manager technically made the final decision, but the AI system may still have materially influenced the outcome.
The same basic issue can arise with hiring, compensation, discipline, or termination.
The EEOC's worker guidance recognizes that AI can be involved in promotion and pay decisions as well as layoffs and termination decisions. EEOC
This is one reason it is misleading to think only about whether a machine "made the decision."
A better approach is to consider the entire decision process:
- What information was collected?
- What information did the system analyze?
- How did the system evaluate or rank the employee?
- Did the system make a recommendation?
- How much did the manager rely on that recommendation?
- Was there an opportunity to correct inaccurate information?
- Did the decision affect hiring, pay, promotion, discipline, or termination?
- Did the outcome potentially discriminate against a protected group?
- Does a state or local law impose additional requirements?
That broader picture can be more important than the simple question of whether a person or computer technically made the final decision.
Federal Employment Protections Still Apply When AI Is Involved
One of the most important points for employees is that AI does not create an exception to existing federal employment-discrimination laws.
The EEOC states that federal employment-discrimination laws protect workers when AI systems are used to discriminate based on protected characteristics including race, color, religion, sex, national origin, age for workers 40 and older, disability, and genetic information. The agency also notes that existing protections can apply to reasonable accommodation issues involving disability, religion, and certain pregnancy-related limitations. EEOC
That means an employer cannot simply avoid employment-discrimination requirements by saying:
"The computer made the decision."
If an AI system contributes to a discriminatory employment outcome, the involvement of technology does not automatically eliminate the underlying legal protections.
Disability Can Be an Important Issue
AI-based workplace evaluations can create particular concerns for employees with disabilities.
For example, an automated system might evaluate speed, communication style, facial expressions, voice characteristics, or other signals that do not accurately reflect an employee's ability to perform essential job duties.
The EEOC specifically recognizes disability-related concerns involving software, algorithms, and artificial intelligence. EEOC
An employee who believes an automated evaluation does not account for a disability-related limitation should not assume that the score is necessarily the end of the matter.
The relevant question may include whether the employer properly considered applicable accommodation obligations and whether the technology is creating a discriminatory barrier.
Workplace AI Can Also Raise Labor-Law Issues
Employment rights are not limited to discrimination laws.
The National Labor Relations Act protects many private-sector employees who act together to improve their wages, benefits, or working conditions, whether or not they belong to a union. The National Labor Relations Board says protected concerted activity can include employees discussing working conditions, raising group complaints, circulating petitions, or acting together concerning workplace issues. National Labor Relations Board
This can matter when an employer uses electronic monitoring or algorithmic management.
The NLRB's current General Counsel guidance states that electronic monitoring and algorithmic management practices can implicate NLRA rights when they tend to interfere with or prevent reasonable employees from engaging in activity protected by the Act. The NLRB also expressly cautions that this material represents the General Counsel's position and has not been reviewed or approved by the Board. National Labor Relations Board
That distinction is important because an older NLRB General Counsel memorandum concerning electronic monitoring and algorithmic management should not be treated as current binding Board law.
For workers, the practical takeaway is simpler: AI monitoring does not erase existing rights concerning protected workplace activity.
For example, if employees are collectively discussing wages or working conditions, an employer cannot necessarily avoid labor-law restrictions simply because monitoring is being performed by software rather than a person.
Most private-sector employees are covered by the NLRA, but there are important exclusions, including certain government employees, agricultural laborers, domestic workers, independent contractors, supervisors, and workers covered by the Railway Labor Act. National Labor Relations Board
There Is No Single Federal "AI at Work" Rule
This is where workplace AI becomes more complicated.
There is not one federal rule saying:
"Employers may use AI under these exact conditions everywhere in the United States."
Instead, multiple areas of law can apply depending on what the employer is doing.
A workplace AI system might therefore be subject to different considerations involving:
- Federal employment-discrimination law
- Disability accommodation requirements
- Labor law
- State employment law
- State artificial-intelligence legislation
- Local workplace-AI requirements
- Employer policies and procedures
- The particular facts surrounding the employment decision
The result is a patchwork rather than one universal rule.
That also means an employee in New York City may have different AI-related protections from an employee performing the same job in another part of the country.
New York City Has Specific Rules for Automated Employment Decision Tools
New York City is an important example of a local jurisdiction that has established specific requirements for certain automated employment decision tools.
Local Law 144 generally prohibits covered employers and employment agencies from using an automated employment decision tool unless specified requirements are met, including a bias audit within the required period, public availability of information about the audit, and required notices to employees or job candidates. The New York City Department of Consumer and Worker Protection began enforcement in July 2023. NYC Home 4
The rules apply to covered automated employment decision tools, so they should not be interpreted as meaning every piece of workplace software is automatically subject to the law.
For an employee or job candidate in New York City, the practical point is that the jurisdiction has requirements specifically addressing certain automated employment decisions. That makes location relevant when evaluating whether an employer's use of AI complies with applicable requirements.
Illinois Has Specific AI Employment Protections
Illinois provides another important example.
The Illinois Human Rights Act contains an employment provision addressing the use of artificial intelligence. The law prohibits covered uses of AI that have the effect of subjecting employees to discrimination based on protected classes under the Act, including in areas such as recruitment, hiring, promotion, renewal of employment, training or apprenticeship selection, discharge, discipline, tenure, and the terms, privileges, or conditions of employment. It also contains a notice requirement concerning covered AI use. Illinois General Assembly
The relevant statutory changes include provisions effective January 1, 2026. Illinois General Assembly
This is a useful example of why employees should not assume that federal law provides the complete answer. A state may impose additional requirements concerning AI in employment.
Colorado Has an AI Framework for Consequential Decisions
Colorado also enacted a significant automated decision-making framework in 2026.
Colorado SB 26-189 addresses automated decision-making technology used in consequential decisions. The law defines covered automated decision-making technology broadly enough to include systems that process personal data and generate outputs such as predictions, recommendations, classifications, rankings, or scores used to make, guide, or assist certain decisions. Employment is included among the areas covered by the definition of a consequential decision. Colorado General Assembly
The law establishes rights and requirements involving areas such as notice, personal-data access and correction, and meaningful human review following certain adverse consequential decisions.
The timing matters.
Colorado's fiscal materials explain that certain provisions concerning deployer disclosures, consumer rights, and Attorney General enforcement take effect upon the law becoming effective, while other provisions take effect January 1, 2027 and apply to covered consequential decisions made on or after that date. Colorado General Assembly
So an employee should not describe the Colorado framework as though every provision is already operating in exactly the same way in October 2026.
California Is Adding Further Worker Protections, But Timing Matters
California provides another example of why the date of a law matters.
California SB 947, addressing automated decision systems in employment, was signed on September 30, 2026. California's governor described the measure as part of a package of new worker protections involving AI and automated workplace technology. The governor's announcement says the legislation includes protections intended to prevent employers from relying only on AI for certain disciplinary or termination decisions. Governor of California
However, the new employment provisions are not simply a set of rules that should be treated as fully operative immediately after enactment. The legislation states that the relevant part becomes operative July 1, 2027. CALege
That distinction is especially important in an article about current employee rights. A law can be enacted today but have important provisions that become operative later.
For workers, this is a good reason to check the current version and effective date of a state law before relying on it in a workplace dispute.
What to Do If You Think an AI Decision Was Wrong
If you believe AI played a role in an unfair employment decision, your first objective should be to preserve information rather than immediately assume why the decision happened.
Start by documenting what you know.
1. Write down the employment decision
Record what happened, when it happened, who communicated it, and what explanation the employer provided.
For example, document whether you were:
- Denied a promotion
- Given a poor performance rating
- Placed on a performance plan
- Disciplined
- Demoted
- Denied a pay increase
- Suspended
- Terminated
Keep the description factual.
2. Preserve relevant communications
Keep copies of communications you are legally and practically permitted to retain, such as notices, performance documents, written policies, emails, and messages concerning the decision.
Do not improperly access confidential company systems or take information you are not authorized to possess.
The goal is to preserve evidence that you already have legitimate access to.
3. Look for information about automated systems
Check employee handbooks, privacy notices, workplace technology policies, onboarding materials, and notices provided by the employer.
You may find information indicating that the company uses monitoring, automated assessments, productivity systems, or other technology.
In jurisdictions with specific notice requirements, those notices can also be relevant.
4. Ask focused questions
If appropriate, ask your employer or HR department questions such as:
- Was an automated system used in this decision?
- What type of information was considered?
- Was the decision reviewed by a manager?
- Was there an opportunity to correct inaccurate information?
- What policy governs the system's use?
You may not receive every answer you request, but asking specific questions can help establish what happened.
5. Compare the decision with your actual work
If an automated score or recommendation appears inaccurate, identify the specific factual problems.
For example, instead of simply saying:
"The AI got it wrong."
you could identify that the system counted a training session as inactive time, failed to account for work performed away from a computer, or relied on information that was factually incorrect.
Specific evidence is generally more useful than a general objection.
Pay Particular Attention to Discrimination or Retaliation
An AI-related employment complaint can involve more than disagreement with an algorithm's accuracy.
Ask whether the outcome appears connected to a protected characteristic or protected workplace activity.
For example, potential issues could arise if an automated system appears to disadvantage employees because of disability, age, sex, race, national origin, religion, or another protected characteristic covered by applicable law. The EEOC specifically recognizes that existing federal employment-discrimination protections can apply when AI systems are used in discriminatory ways. EEOC
Labor-law issues can also arise when workplace monitoring is connected to protected concerted activity.
The NLRB explains that employees can have protected rights to act together regarding workplace conditions, including wages, benefits, and other terms of employment. National Labor Relations Board
That does not mean every complaint about workplace monitoring is protected activity. The legal protections depend on the circumstances and applicable law.
Keep an AI Employment Decision in Perspective
AI can be powerful, but an AI-generated score or recommendation is not necessarily an objective measurement of an employee's value.
An automated system depends on the data it receives, the way it was designed, the factors selected by its developer or employer, and the way people interpret its output.
A system can also be technically accurate at measuring one narrow variable while still being a poor way to evaluate the broader question an employer actually cares about.
For example, measuring keyboard activity may tell an employer something about keyboard activity. It does not automatically tell the employer everything about an employee's judgment, teamwork, customer service, problem-solving, or contribution to the business.
This distinction matters because employment decisions often involve qualities that are difficult to reduce to a single score.
For employees, the most useful mindset is therefore neither "AI is always wrong" nor "AI is objective."
Instead, ask:
What exactly is the system measuring, and how is that information being used?
What Employees Should Watch for Going Forward
Workplace AI rules are continuing to develop, and the legal landscape can change as states and cities adopt new requirements.
Employees can protect themselves by becoming familiar with the technology their workplace uses.
Pay attention when your employer introduces:
- New productivity-monitoring software
- Automated performance scores
- AI-based hiring or promotion tools
- Employee-location tracking
- Automated scheduling or assignment systems
- New workplace surveillance technology
- AI-based disciplinary systems
- Automated recommendations affecting pay or advancement
You do not need to assume that every new system violates your rights.
Instead, understand what it does.
If the employer explains the system, read the applicable policy. If a decision appears inaccurate, document the problem. If the outcome appears discriminatory or retaliatory, preserve relevant evidence and consider whether a government agency, worker representative, or employment attorney can help you evaluate the situation.
A Practical Way to Think About AI Workplace Rights
The easiest way to evaluate an AI-related workplace problem is to work through four questions.
This framework prevents a common mistake: treating "AI" as the legal issue by itself.
Usually, the more important question is what the employer is doing with AI.
An AI system used to organize information for a manager is not necessarily equivalent to an automated system that materially influences a termination. Likewise, a workplace monitoring tool may raise different questions from an AI hiring system.
The facts matter.
The Bottom Line for U.S. Employees
Your employer may be able to use AI to monitor, evaluate, or assist with employment decisions, but that does not give the employer unlimited freedom to discriminate, retaliate, or ignore other applicable workplace protections.
Federal employment laws can apply when AI contributes to discriminatory employment decisions. Labor-law protections can also matter when electronic monitoring or algorithmic management interferes with protected employee activity. State and local governments are adding their own requirements, including rules concerning notice, bias audits, data rights, and human involvement in certain automated decisions. EEOC
The most important distinction is between AI being used as a workplace tool and AI determining or materially influencing what happens to an employee. Those situations can involve different legal considerations.
If you are affected by an AI-related workplace decision, document what happened, preserve relevant information you legitimately have access to, identify how the system may have influenced the decision, and check the laws that apply where you work.
Conclusion
AI can now be part of many stages of employment, from hiring and performance evaluation to workplace monitoring, pay, discipline, and termination. But there is no single nationwide rule that makes every workplace use of AI legal or illegal.
For employees, the key is to look beyond the word "AI." Determine what the technology is actually doing, what information it uses, how much influence it has over the employment decision, and whether federal, state, or local protections apply.
If an AI-driven outcome appears inaccurate, discriminatory, or connected to protected workplace activity, do not rely only on the automated score or recommendation. Document the decision, preserve relevant information, ask focused questions, and consider the appropriate government or legal resources for your situation.
As workplace AI rules continue to develop, knowing how the technology is being used and which protections apply is likely to be more useful than assuming that every AI decision follows one universal set of rules.
Frequently asked questions
Does my employer have to tell me if AI is being used to monitor or evaluate me?
There is no single federal rule requiring every employer to disclose every use of AI for workplace monitoring or evaluation. Notice requirements can depend on the type of AI, the employment decision, and the state or local law that applies. Some jurisdictions have specific disclosure requirements.
Can my employer use AI to track my productivity?
An employer may use technology, including AI, to monitor or analyze productivity, computer activity, communications, location, work patterns, or other workplace information. Whether the monitoring is lawful depends on the circumstances, applicable employment laws, company policies, and state or local requirements.
Can an employer fire me because of an AI decision?
AI use does not automatically make a termination unlawful, but an employer generally cannot use AI as a way to carry out an otherwise unlawful employment decision. Federal employment-discrimination laws can apply when AI contributes to discrimination based on protected characteristics. State and local laws may provide additional protections.
Does a human have to review an AI employment decision?
There is no universal federal requirement that a human review every AI-assisted employment decision. Human-review requirements can depend on the jurisdiction and type of automated decision. For example, Colorado's enacted AI law includes meaningful human-review rights for certain covered consequential decisions beginning in 2027.
Can AI used by an employer discriminate against employees?
Yes. AI systems can contribute to discriminatory employment outcomes, including when the system uses biased data, evaluates characteristics inappropriately, or creates barriers for people with disabilities. Federal employment-discrimination laws can apply when AI is used in employment decisions in a discriminatory way.
What should I do if I think AI was used unfairly against me at work?
Keep records of the decision, relevant communications, performance information, policies, notices, and other evidence you have access to. Ask appropriate questions about how the decision was made and whether AI or an automated system was involved. If you believe discrimination or another legal violation occurred, consider contacting the appropriate government agency or qualified employment attorney.
Can my employer monitor my email, keystrokes, or computer activity with AI?
1. AI can be used to analyze workplace information such as email, keystrokes, cursor activity, computer use, task time, location, and other activity. The legality and disclosure requirements depend on the circumstances and applicable laws. Workplace monitoring can also raise additional issues when it interferes with legally protected employee activity.