Does employee monitoring hurt productivity?

Updated on 6 July 2026

It can. In 2024, Cornell ILR School researchers found that people monitored by AI for evaluation performed worse, complained more and were likelier to intend quitting than people monitored by humans, but when the same monitoring was framed as developmental feedback those effects largely disappeared. How monitoring is used decides its effect.

What does the research say about monitoring and productivity?

That the damage comes from how monitoring is used, not from measurement itself. In studies from Cornell University's ILR School, published in 2024, participants monitored by AI reported a greater loss of autonomy than those overseen by humans, generated fewer ideas, complained more and were more inclined to quit. In the final study, participants told that the analysis of their work would provide developmental feedback, rather than evaluate their performance, no longer perceived the loss of autonomy and reported no greater intention to quit. A separate 2022 meta-analysis found electronic monitoring slightly lowers job satisfaction and slightly raises stress.

Why does evaluative monitoring backfire?

Because being judged through proxies changes behaviour before it changes output. When people feel watched for a verdict, autonomy drops, resistance rises, and effort migrates from the work to the metric. Input-based activity scores make this worse: they are easy to game, which is why trackers now ship fake-activity detection (Hubstaff's Insights add-on and Time Doctor's Unusual Activity Report both exist to catch manufactured input). A workplace where people optimise for looking busy has paid the productivity cost twice.

What does developmental monitoring look like in practice?

It gives the data to the person it is about, as feedback rather than a verdict. ScreenJournal is an AI work visibility tool that reads on-screen work as it happens, turns it into a detailed timeline of what each person actually did, and then deletes the raw screen data. Timelines accumulate into a searchable chronicle of everyone's work history, and from them ScreenJournal generates timesheets and reports automatically and drafts standup summaries on request, answering questions about any of it in plain English.

The developmental defaults are built in. Employees see the same activity view managers do, any score can be contested through "Change my score", nudges are off by default, and entries can be redacted before a manager sees them. The work timeline is the shared record both sides talk from. For how this differs from productivity-scoring tools, see ScreenJournal vs ActivTrak.

One honest nuance: much of this evidence comes from controlled studies and scenario experiments rather than years of field data, so treat it as a strong signal about design, not a precise forecast for your team.

Frequently asked questions

Is being monitored by AI worse than being monitored by a person?

In the Cornell studies it was. People monitored by AI felt a greater loss of autonomy and showed more resistance than people monitored by humans. Framing the AI monitoring as developmental feedback removed that difference.

Does employee monitoring make people want to quit?

It can. Cornell participants subject to evaluative AI monitoring reported greater intention to quit, and meta-analytic evidence links electronic monitoring to slightly higher stress and slightly lower job satisfaction.