AI is moving from assistant to workplace evaluator. Firms are increasingly using it to monitor or assess employees, creating pressure for workers to demonstrate skills they may not possess with confidence.
That pressure has produced a behavior called performative AI: workers exaggerate how much they use AI or how much they understand it. The technology may be sold as a productivity tool, but the workplace response can become a performance of competence—because apparently ordinary work now needs an audition for the algorithm.
When AI monitoring becomes part of the job
Instructors at tech training company Multiverse recently blew the whistle on an AI monitoring system they described as “remorseless” and “unnerving.” The description points to a system that does more than support work; it watches and assesses the people expected to use it.
The Co-Op has also used automated listening technology that rates every phone call some of its workers make to customers. That turns each conversation into an assessed event, with an automated system judging calls as part of the work itself.
These examples show two sides of the same workplace shift. AI can monitor activity inside a training company and rate customer calls at a retailer, placing workers under assessment even when their main duties differ.
The issue is not only whether these systems monitor employees. It is also what the monitoring encourages employees to do. When workers believe they must prove AI expertise, showing confidence can matter as much as having a useful result.
Pressure creates performative AI
Almost half of US employees surveyed by workforce intelligence AI company Visier said they had exaggerated either how much they use AI or their expertise with it. The finding describes a gap between the pressure to appear capable and the reality of how confident workers feel.
Visier found that 45% of workers felt pressure to leverage AI even when they were not confident about how to do so effectively. That figure puts a number on the tension: employees are being pushed toward AI use while many remain unsure how to make that use work.
Andrea Derler, principal of research and value at Visier, said performative AI happens because of pressure for workers to show their expertise. Her explanation places the cause with workplace expectations, not with a sudden desire among employees to inflate their skills.
That distinction matters. If workers exaggerate their AI use because they feel watched or judged, the behavior becomes a response to the system around them rather than a simple question of honesty.
Automated monitoring can therefore create an awkward loop. Firms use AI to assess employees, employees feel pressure to demonstrate AI expertise, and some respond by overstating their use or ability.
The loop also changes what competence looks like at work. A worker can be judged not only by what they produce, but by whether they appear to use AI in the expected way—even when they lack confidence in doing so effectively.
That is the more important story behind performative AI. The technology is not only measuring work; it is shaping how workers present themselves inside the workplace.
Companies now have two facts to confront: AI monitoring is expanding, and almost half of surveyed US employees say they have exaggerated their AI use or expertise. Treating the second fact as a worker problem misses the pressure described by the first.
AI may be good at rating calls and assessing employees. It has yet to prove that making workers perform confidence produces better work.
Based on
