51% of US workers say their employer uses technology to monitor them while they work, and 56% of those monitored workers typically feel tense or stressed at work, compared with 40% of workers who are not monitored (American Psychological Association, Work in America 2023). The research base points in the same direction: a meta-analysis of 94 samples covering 23,461 workers found no evidence that electronic performance monitoring improves performance, but did find it associated with higher stress (Ravid et al., Personnel Psychology). Monitoring is not only a US story either, since 25% of European workers report digital tools supervising their work and behaviour and 48% say technology sets the pace of their work (EU-OSHA, OSH Pulse 2025). The legal ground is shifting at the same time, with US federal guidance withdrawn in 2025 while the EU’s ban on workplace emotion recognition took effect. We aggregated data from the American Psychological Association, Pew Research Center, EU-OSHA, the US Government Accountability Office, peer-reviewed meta-analyses, state statutes, the European Commission, Microsoft’s Work Trend Index and named vendor surveys, each listed in the methodology. Attendance tracking through badge swipes is covered separately in our hybrid work mandate compliance statistics.
TL;DR
- 51% of US workers say their employer uses technology to monitor them at work (APA, Work in America 2023).
- 56% of monitored workers feel tense or stressed at work vs 40% of unmonitored workers (APA, Work in America 2023).
- 42% of monitored workers intend to look for a new job within 12 months vs 23% of unmonitored workers (APA, Work in America 2023).
- 25% of European workers say digital technologies supervise or monitor their work and behaviour (EU-OSHA, OSH Pulse 2025).
- 80% of employers surveyed use employee monitoring software, and 57% adopted it within the previous six months (ExpressVPN vendor survey 2021).
- 61% of US adults oppose employers using AI to track workers’ movements, and 51% oppose recording what workers do on their computers (Pew Research Center 2023).
- 81% of US adults say AI monitoring would make workers feel inappropriately watched (Pew Research Center 2023).
- Electronic monitoring shows no relationship with performance (r = -0.01) across 70 samples (Siegel, Konig and Lazar, Computers in Human Behavior Reports 2022).
- Only 12% of leaders have full confidence their team is productive, while 87% of employees say they are (Microsoft Work Trend Index 2022).
- Focus efficiency in monitoring-software data fell to 60%, a three-year low (ActivTrak, 2026 State of the Workplace).
- NLRB memo GC 23-02 on electronic monitoring was rescinded on February 14, 2025 (NLRB GC 25-05).
- The EU AI Act ban on workplace emotion recognition has applied since 2 February 2025 (European Commission).
1. How Many Workers Are Monitored
The headline number depends on who is asked and what counts as monitoring. In APA’s 2023 survey, 51% of US workers said their employer uses technology such as computers, software, cameras or bar-code scanners to monitor them while working, with almost no difference between office workers (49%) and manual laborers (49%), and a slightly higher rate in customer, client and patient services (55%) (APA, 2023 Work in America: AI and monitoring). Surveillance has stopped being a warehouse-and-call-center issue; desk workers report it at the same rate.
Europe’s newest data put a narrower lens on the same trend. In EU-OSHA’s OSH Pulse 2025, fielded in March and April 2025 with 28,220 workers in 30 countries, 25% said digital technologies are used to supervise or monitor their work and behaviour, while 27% reported technology that automatically allocates tasks, working time or shifts (EU-OSHA, OSH Pulse 2025 summary). The lower European share likely reflects both stricter data protection rules and a narrower question wording, so the two figures should not be compared directly.
| Metric | Value | Source |
|---|---|---|
| US workers who say their employer monitors them with technology | 51% | APA, Work in America 2023 (n = 2,515) |
| Monitoring rate: office work / manual labor / customer, client or patient services | 49% / 49% / 55% | APA, Work in America 2023 |
| European workers whose work and behaviour is supervised or monitored by digital tech | 25% | EU-OSHA, OSH Pulse 2025 (n = 28,220) |
| Tasks, working time or shifts allocated automatically | 27% | EU-OSHA, OSH Pulse 2025 |
| Performance ratings by third parties via digital tech | 26% | EU-OSHA, OSH Pulse 2025 |
| Automated instructions or directions to complete work | 26% | EU-OSHA, OSH Pulse 2025 |
| Monitoring of heart rate, blood pressure or posture | 7% | EU-OSHA, OSH Pulse 2025 |
| Employers using employee monitoring software (vendor survey) | 80% (57% adopted in the prior six months) | ExpressVPN survey 2021 (2,000 employers) |
Context note: the APA question measures what workers know about, so covert monitoring is not captured and the true share may be higher. The ExpressVPN figure comes from a VPN vendor’s own survey of 2,000 employers and 2,000 employees in early 2021, during the remote-work surge; it is labelled as a vendor survey and is the most recent available data of its kind that we could verify. A Gartner estimate widely quoted on this topic could not be loaded from a Gartner primary page and was excluded.
2. What Gets Tracked: Keystrokes, Screens, Apps and Communications
Monitoring suites rarely stop at one signal. Among employers using monitoring software in ExpressVPN’s survey, 94% monitored email, 87% calls, 87% video and 85% messages, and two-thirds tracked websites visited and time spent on them (ExpressVPN, Surveillance on the remote workforce, 2021). Real-time screen monitoring (53%) and periodic screen capture (33%) show how far tooling has moved beyond time sheets.
US regulators described the same inventory. The CFPB’s Circular 2024-06, issued October 24, 2024, listed employer monitoring of sales interactions, driving habits, task completion time, message frequency, meeting duration, web browsing via screenshots and keystroke frequency as data that can feed third-party worker dossiers (CFPB, Circular 2024-06). GAO’s 2025 review summarized “bossware” as cameras, microphones, computer monitoring software and geolocation trackers.
| Metric | Value | Source |
|---|---|---|
| Employers monitoring email / calls / video / messages | 94% / 87% / 87% / 85% | ExpressVPN survey 2021 |
| Websites visited and time spent on them | 66% | ExpressVPN survey 2021 |
| Real-time screen monitoring / apps used | 53% / 53% | ExpressVPN survey 2021 |
| Active work hours and log times tracked | 46% | ExpressVPN survey 2021 |
| Productive vs unproductive hours logged | 31% | ExpressVPN survey 2021 |
| Periodic screen capture | 33% | ExpressVPN survey 2021 |
| Data types named in federal guidance | Keystroke frequency, browsing screenshots, task time, message frequency, meeting duration, driving habits | CFPB Circular 2024-06 (2024, withdrawn 2025) |
Context note: keystroke logging as a stand-alone feature was not broken out in the ExpressVPN results, so we do not report a keystroke-specific adoption rate. The CFPB circular was withdrawn in May 2025 (see section 7), but its description of the data employers collect remains a useful federal inventory.
3. What Time-Tracking Data Shows About the Workday
Activity-tracking platforms now publish their own aggregate telemetry, which shows what the software sees rather than what workers report. ActivTrak’s 2026 State of the Workplace, built from 443+ million hours logged by 163,638 employees at 1,111 companies between 2023 and 2025, found focus efficiency fell to 60%, a three-year low, while the average focused session shrank to 13 minutes 7 seconds (ActivTrak, 2026 State of the Workplace). Collaboration time rose 34% to 52 minutes a day and multitasking climbed 12% to 1 hour 33 minutes.
The striking detail is how productive the dashboards look on paper. Productive hours rose 5% to 6 hours 36 minutes a day inside an 8 hour 44 minute workday, yet the share of uninterrupted deep work is shrinking. That is the measurement problem at the heart of keystroke and activity tracking: it counts motion more reliably than it counts thinking. Interruption patterns that drive this are explored in our open-plan office distraction statistics.
| Metric | Value | Source |
|---|---|---|
| Dataset | 1,111 companies, 163,638 employees, 443+ million hours (2023-2025) | ActivTrak, 2026 State of the Workplace |
| Productive hours per day | 6 hours 36 minutes (up 5%) | ActivTrak, 2026 State of the Workplace |
| Average workday length | 8 hours 44 minutes (down 2%) | ActivTrak, 2026 State of the Workplace |
| Focus efficiency | 60% (three-year low) | ActivTrak, 2026 State of the Workplace |
| Average focused session | 13 minutes 7 seconds (down 9%) | ActivTrak, 2026 State of the Workplace |
| Daily collaboration / multitasking time | 52 minutes (up 34%) / 1 hour 33 minutes (up 12%) | ActivTrak, 2026 State of the Workplace |
| Saturday work time | 4 hours 37 minutes (up 46%) | ActivTrak, 2026 State of the Workplace |
| Employees using AI tools | 80% (up from 53%) | ActivTrak, 2026 State of the Workplace |
Context note: ActivTrak sells workforce analytics software, and these are aggregates from its own customer base, which is not a random sample of employers. Definitions of “productive” and “focused” time are set by the vendor and by each customer’s app classifications.
4. Stress, Wellbeing and Turnover Among Monitored Workers
The mental health gap is consistent across every APA measure. Monitored workers were more likely than unmonitored workers to feel tense or stressed at work (56% vs 40%), to say work has a negative effect on their mental health (45% vs 29%) and to report poor or fair mental health (32% vs 24%) (APA, Electronically monitoring your employees? It’s impacting their mental health, 2023). They also felt micromanaged far more often (51% vs 33%), which is the mechanism employees describe.
The retention signal matters most for employers: 42% of monitored workers planned to look for a new job within a year, against 23% of unmonitored workers. European data show the pressure side of the same equation, with 48% of workers saying digital technologies determine the speed or pace of their work and 28% saying they increase workload (EU-OSHA, OSH Pulse 2025). Sectors where every minute is already measured show where this leads, as covered in our call center agent burnout and turnover statistics.
| Metric | Value | Source |
|---|---|---|
| Typically feel tense or stressed at work: monitored vs unmonitored | 56% vs 40% | APA, Work in America 2023 |
| Work has a negative impact on mental health | 45% vs 29% | APA, Work in America 2023 |
| Poor or fair mental health | 32% vs 24% | APA, Work in America 2023 |
| Feel micromanaged | 51% vs 33% | APA, Work in America 2023 |
| Intend to seek a new job in the next 12 months | 42% vs 23% | APA, Work in America 2023 |
| Worry employer uses technology to spy on them | 47% vs 22% (34% of all workers) | APA, Work in America 2023 |
| Digital technologies determine speed or pace of work | 48% | EU-OSHA, OSH Pulse 2025 |
| Digital technologies increase workload / result in working in isolation | 28% / 30% | EU-OSHA, OSH Pulse 2025 |
Context note: the APA comparisons are cross-sectional, so they show association rather than proof that monitoring causes the gap; monitored jobs may differ in other ways. APA’s survey was conducted by The Harris Poll from April 17 to 27, 2023, with a margin of error of plus or minus 3.1 points. Most recent available data: APA, 2023.
5. The Trust Gap: Productivity Paranoia and Public Opinion
Monitoring is often a response to anxiety on the employer side rather than evidence of shirking. Only 12% of leaders said they had full confidence their team is productive, while 87% of employees said they are productive at work (Microsoft Work Trend Index, Hybrid Work Is Just Work, 2022). Microsoft’s survey of 20,006 workers in 11 countries labelled the result “productivity paranoia” and warned that tracking activity instead of impact can produce “productivity theater”. In ExpressVPN’s vendor survey, 59% of employers said they do not trust employees to work without digital supervision, and 49% of employees admitted pretending to be online while doing non-work activities.
The public sides with the watched. 61% of US adults oppose employers using AI to track workers’ movements while they work, 56% oppose tracking when office workers are at their desks and 51% oppose recording exactly what workers do on their computers (Pew Research Center, AI in hiring and evaluating workers, 2023). About eight in ten (52% definitely, 29% probably) said workers would feel inappropriately watched, and 66% expected the collected information to be misused.
| Metric | Value | Source |
|---|---|---|
| Leaders with full confidence their team is productive | 12% | Microsoft Work Trend Index 2022 (n = 20,006) |
| Employees who say they are productive at work | 87% | Microsoft Work Trend Index 2022 |
| Leaders who say hybrid work makes confidence in productivity challenging | 85% | Microsoft Work Trend Index 2022 |
| Employers who do not trust employees without digital supervision | 59% | ExpressVPN survey 2021 |
| Employees who pretended to be online while doing non-work activities | 49% | ExpressVPN survey 2021 |
| US adults opposing AI tracking of movements / desk time / computer activity | 61% / 56% / 51% | Pew Research Center 2023 (n = 11,004) |
| Full-time workers opposing AI tracking of movements | 68% | Pew Research Center 2023 |
| Say AI monitoring would make workers feel inappropriately watched | 81% (52% definitely, 29% probably) | Pew Research Center 2023 |
Context note: Pew’s survey was fielded December 12-18, 2022, and published in April 2023; it is the most recent available Pew data on this question. Driving-behavior monitoring was the only use with more support (43%) than opposition (34%), and 70% opposed AI analysis of workers’ facial expressions. Microsoft’s survey ran July 7 to August 2, 2022 (most recent available data on this measure).
6. Does Monitoring Work? What the Research Says
The academic verdict is unusually consistent. A meta-analysis of 94 independent samples covering 23,461 workers found “no evidence” that electronic performance monitoring improves worker performance, while its presence was associated with increased stress regardless of how monitoring was done (Ravid, White, Tomczak, Miles and Behrend, Personnel Psychology). Transparent and less invasive monitoring was linked to more positive worker attitudes, which suggests the design of a program matters more than its existence. A second meta-analysis of 70 samples found monitoring correlated with lower job satisfaction (r = -0.10), higher stress (r = .11), no change in performance (r = -0.01) and slightly more counterproductive behavior (r = 0.09) (Siegel, Konig and Lazar, Computers in Human Behavior Reports 2022).
Newer studies add nuance on AI-run surveillance. Across four experiments with up to 814 participants, people monitored by an algorithm rather than a human felt less autonomy, performed worse on idea-generation tasks and reported more intent to resist, although framing the monitoring as developmental reduced the effect (Schlund and Zitek, Communications Psychology 2024). A smaller two-wave study of 223 Chinese employees found developmental and preventive monitoring positively related to job performance through self-efficacy (Zhang, Sun and Jin, Behavioral Sciences 2025). GAO’s 2025 report, which assessed 122 studies published from 2020 to 2024 and interviewed 11 stakeholder organizations, concluded that effects on stress and anxiety vary with employer transparency (GAO-25-107126).
| Metric | Value | Source |
|---|---|---|
| Effect of electronic monitoring on performance (94 samples, N = 23,461) | No evidence of improvement; associated with higher stress | Ravid et al., Personnel Psychology (2022 online, vol. 76) |
| Correlation with job satisfaction / stress (70 samples, 233 effect sizes) | r = -0.10 / r = .11 | Siegel, Konig and Lazar, CHB Reports 2022 |
| Correlation with performance / counterproductive work behavior | r = -0.01 / r = 0.09 | Siegel, Konig and Lazar, CHB Reports 2022 |
| Algorithmic vs human surveillance (4 experiments, largest N = 814) | Less autonomy, worse performance, more intent to resist | Schlund and Zitek, Communications Psychology 2024 |
| AI surveillance and wellbeing, Chinese manufacturing managers (n = 482, three waves) | Surveillance lowered wellbeing; resilience strengthened the surveillance-distress link | Chu and Chu, Frontiers in Psychology 2026 |
| Perceived algorithmic surveillance among crowdworkers (n = 435, three waves) | Lower trust and fairness, higher privacy concern | van Zoonen et al., Human Relations 2025 |
| Developmental monitoring and performance (n = 223) | Positive effect via self-efficacy | Zhang, Sun and Jin, Behavioral Sciences 2025 |
| Studies assessed by GAO on digital surveillance effects | 122 (published 2020-2024) | GAO-25-107126 (2025) |
Context note: the two meta-analyses draw on overlapping literatures and mostly pre-2020 studies, so they are the most recent available syntheses rather than a measure of today’s AI-driven tools. The positive findings come from single-country samples with self-reported outcomes. Monitoring that extends to camera and audio feeds also overlaps with the strain documented in our video meeting fatigue statistics.
7. Law and Regulation: Notice Rules, Rescissions and the EU Ban
Disclosure is the core legal obligation in the US, and penalties are small. New York’s Civil Rights Law section 52-c requires employers that monitor phone, email or internet use to give prior written notice upon hiring, obtain an acknowledgment and post a notice, with civil penalties of up to 500 dollars for a first offense, 1,000 dollars for a second and 3,000 dollars for the third and later offenses (New York State Senate, CVR 52-C*2). Delaware, which legislated in 2001, allows either a daily electronic notice or a one-time acknowledged notice and sets a 100 dollar penalty per violation (Delaware Code, Title 19, section 705). In Canada, Ontario requires employers with 25 or more employees on January 1 to have a written electronic monitoring policy in place before March 1 and give copies within 30 calendar days (Government of Ontario, Employment Standards Act guide).
Federal guidance moved in the opposite direction. NLRB General Counsel memo GC 23-02, announced October 31, 2022, proposed that surveillance which would tend to interfere with protected activity is presumptively unlawful, but GC 25-05 rescinded it on February 14, 2025 (NLRB, GC 25-05). The CFPB’s worker-dossier circular was withdrawn in a notice published May 12, 2025 (Federal Register 2025-08286). Europe went the other way: the AI Act’s ban on emotion recognition in workplaces took effect on 2 February 2025 (European Commission, AI Act).
| Metric | Value | Source |
|---|---|---|
| New York penalties: first / second / third and later offenses | Up to 500 / 1,000 / 3,000 dollars | NY Civil Rights Law section 52-c |
| Delaware penalty per violation | 100 dollars (enacted 2001) | Delaware Code Title 19, section 705 |
| Ontario policy threshold and deadlines | 25+ employees on January 1; policy before March 1; copies within 30 days (first deadline October 11, 2022) | Government of Ontario, ESA guide |
| NLRB electronic monitoring memo | GC 23-02 issued October 31, 2022; rescinded February 14, 2025 | NLRB, GC 25-05 |
| CFPB Circular 2024-06 on worker dossiers and scores | Issued October 24, 2024; withdrawn May 12, 2025 | CFPB; Federal Register 2025-08286 |
| EU AI Act ban on workplace emotion recognition | In force from 2 February 2025 | European Commission, AI Act |
| EU high-risk obligations for employment and worker-management AI | From 2 December 2027 (AI Omnibus timeline) | European Commission, AI Act |
| CNIL fine on an employee monitoring software developer | 20,000 euros (lack of cooperation, 2023) | CNIL, sanctions list |
Context note: GAO found that federal agencies that had previously addressed digital surveillance “have either rescinded these prior efforts or are reassessing” them in 2025 (GAO-25-107126). The CNIL entry appears on its sanctions list without naming the company. Connecticut also has an electronic monitoring notice statute (General Statutes section 31-48d), but its full text could not be loaded, so its penalties are not reported here.
Summary: Employee Monitoring Software by the Numbers
| Metric | Value | Source |
|---|---|---|
| US workers who say their employer monitors them | 51% | APA, Work in America 2023 |
| Monitored vs unmonitored workers feeling tense or stressed | 56% vs 40% | APA, Work in America 2023 |
| Monitored vs unmonitored workers planning to job hunt | 42% vs 23% | APA, Work in America 2023 |
| Monitored vs unmonitored workers feeling micromanaged | 51% vs 33% | APA, Work in America 2023 |
| European workers whose work and behaviour is digitally monitored | 25% | EU-OSHA, OSH Pulse 2025 |
| European workers whose work pace is set by digital tech | 48% | EU-OSHA, OSH Pulse 2025 |
| Employers using monitoring software (vendor survey) | 80% | ExpressVPN survey 2021 |
| Employers monitoring email | 94% | ExpressVPN survey 2021 |
| Employers using real-time screen monitoring | 53% | ExpressVPN survey 2021 |
| Focus efficiency in activity-tracking data | 60% (three-year low) | ActivTrak, 2026 State of the Workplace |
| Average focused session | 13 minutes 7 seconds | ActivTrak, 2026 State of the Workplace |
| Leaders fully confident their team is productive | 12% | Microsoft Work Trend Index 2022 |
| US adults opposing AI tracking of workers’ movements | 61% | Pew Research Center 2023 |
| US adults who say AI monitoring makes workers feel inappropriately watched | 81% | Pew Research Center 2023 |
| Monitoring effect on performance (94 samples) | No evidence of improvement | Ravid et al., Personnel Psychology |
| Monitoring correlation with stress (70 samples) | r = .11 | Siegel, Konig and Lazar, CHB Reports 2022 |
| Studies reviewed by GAO on surveillance effects | 122 | GAO-25-107126 (2025) |
| New York maximum penalty, third offense | 3,000 dollars | NY Civil Rights Law section 52-c |
| NLRB monitoring memo GC 23-02 | Rescinded February 14, 2025 | NLRB GC 25-05 |
| EU workplace emotion recognition ban | Since 2 February 2025 | European Commission, AI Act |
Methodology and Sources
Every figure above was read during research for this article on the publisher’s own page, in a statute, a government document, a peer-reviewed abstract or a named organization’s survey. Vendor surveys and vendor telemetry are labelled as such and were not merged with independent surveys.
- American Psychological Association: 2023 Work in America, AI and monitoring, Electronically monitoring your employees? It’s impacting their mental health (September 7, 2023)
- EU-OSHA: OSH Pulse 2025 summary (fieldwork March 31 to April 14, 2025)
- Pew Research Center: Americans’ views on use of AI to monitor and evaluate workers (April 20, 2023)
- Microsoft: Work Trend Index, Hybrid Work Is Just Work (2022)
- ExpressVPN (vendor survey): Surveillance on the remote workforce (2021)
- ActivTrak (vendor telemetry): 2026 State of the Workplace
- Peer-reviewed research: Ravid et al., Personnel Psychology, Siegel, Konig and Lazar, Computers in Human Behavior Reports 2022, Schlund and Zitek, Communications Psychology 2024; Chu and Chu (Frontiers in Psychology 2026), van Zoonen et al. (Human Relations 2025) and Zhang, Sun and Jin (Behavioral Sciences 2025) were read via Europe PMC and OpenAlex abstracts
- US Government Accountability Office: GAO-25-107126, Digital Surveillance: Potential Effects on Workers and Roles of Federal Agencies (2025)
- Consumer Financial Protection Bureau: Circular 2024-06; withdrawal in Federal Register 2025-08286
- National Labor Relations Board: GC 25-05, Rescission of Certain General Counsel Memoranda (February 14, 2025)
- Statutes and official guidance: New York Civil Rights Law section 52-c, Delaware Code Title 19, section 705, Ontario Employment Standards Act guide, European Commission, AI Act regulatory framework, CNIL sanctions list
- Data watch: Prevalence estimates conflict because they measure different things: APA asks US workers what they know (51%), EU-OSHA asks European workers about a narrower supervision item (25%), and ExpressVPN asks employers that it surveyed during the 2021 remote-work surge (80%). The APA, Pew and Microsoft data are from 2022-2023 and ExpressVPN from 2021, the most recent available verified data on those measures; the APA 2024 Work in America report page we read contained no monitoring figures. ActivTrak’s figures come from its own customers. No employee monitoring market-size estimate is included because the research-firm pages we tried (Grand View Research, MarketsandMarkets, Fortune Business Insights) were blocked or returned a different report, so no cross-check between two firms was possible. Gartner press releases, the UK ICO’s 2023 monitoring guidance announcement and the CNIL announcement of a fine against Amazon France Logistique could not be loaded and were excluded, as were Connecticut’s penalty amounts. The AI Act high-risk dates reflect the European Commission’s page after the AI Omnibus agreement and may change if the final text differs. The Ravid et al. article appeared online in 2022 and in Personnel Psychology volume 76.
Last updated: October 3, 2026. We update this roundup quarterly, and the next refresh is expected when APA publishes its 2027 Work in America survey and EU-OSHA releases the full country-level OSH Pulse 2025 digitalisation data.