Abstract
Orientation: Artificial intelligence (AI) monitoring differs from earlier electronic performance monitoring in three respects: It observes continuously, it infers states that employees never disclose, and it removes human discretion from evaluation. These features change how employees experience oversight.
Research purpose: We examine how perceived AI monitoring relates to turnover intention (TI) through two sequential mechanisms: Perceived privacy intrusion (PPI) and psychological reactance.
Motivation for the study: Monitoring research disagrees about whether oversight harms employees, yet it rarely specifies the psychological route from algorithmic oversight to exit cognition. We address that route directly rather than restating the debate.
Research approach/design and method: We collected cross-sectional survey data from 390 Generation Z employees in Indonesian organisations whose work is subject to automated monitoring. We tested the model using partial least squares structural equation modelling (PLS-SEM).
Main findings: Artificial intelligence monitoring relates most strongly to PPI. Intrusion predicts psychological reactance, and reactance is the most proximal predictor of TI. The central mechanism is a serial cascade: Monitoring raises intrusion, intrusion activates reactance, and reactance drives the intention to leave.
Practical/managerial implications: Because reactance, not intrusion, is the proximal driver of exit, technical privacy fixes alone are insufficient. Restoring discretion, voice and procedural justification addresses the mechanism that the data identify as decisive.
Contribution/value-add: We specify a boundary–autonomy cascade. The contribution is the ordering of appraisal before motivation and before behaviour, not the integration of established theories as such. This ordering reframes turnover as an autonomy-restoration response to oversight that employees judge illegitimate.
Keywords: AI monitoring; privacy intrusion; psychological reactance; turnover intention; algorithmic management; workplace surveillance.
Introduction
Workplace monitoring research carries an unresolved disagreement. One line of work treats oversight as a control that protects performance; another treats it as a stressor that erodes the conditions performance depends on. Decades of evidence show that monitoring can raise stress, reduce perceived control and lower satisfaction when organisations apply it without procedural safeguards (Carayon, 1994; Siegel et al., 2022). The disagreement persists because most studies estimate whether monitoring affects outcomes rather than how it does so. Artificial intelligence (AI)-enabled monitoring now spreads through organisations faster than the disagreement resolves. Industry research and qualitative accounts describe systems that scan internal communication and behavioural data in real time and highlight worker concerns about privacy, autonomy and well-being (Cardon et al., 2023; Sum et al., 2025). Syntheses of AI in organisations similarly document a divide between promised efficiency gains and risks to worker experience and job quality (Bankins et al., 2024). These sources motivate the question but do not answer it. The unanswered question is mechanistic: Through what psychological route does AI monitoring translate into the intention to leave?
Artificial intelligence monitoring is not a faster version of electronic performance monitoring (EPM); it differs in ways that change employee psychology. Four features matter. Opacity hides the evaluative standard, so employees cannot predict what behaviour the system rewards or penalises. Inference extends evaluation beyond observed action to imputed states such as sentiment, intent or risk, so the system generates information that employees never disclosed. Automation removes human discretion, so there is no person to whom an employee can appeal or explain. Permanence extends the temporal reach of evaluation, so past data remain available indefinitely. Each feature produces a distinct psychological consequence: Opacity produces evaluative uncertainty, inference produces informational exposure, automation removes the relational buffer that once softened oversight, and permanence dissolves the sense that evaluation is bounded in time. Electronic performance monitoring research, which studies episodic and observable monitoring, does not capture these consequences (Kalischko & Riedl, 2021; Ravid et al., 2023). The gap is therefore not the absence of monitoring research but the absence of a model that connects these specific features to withdrawal cognition.
We propose a single integrating idea that unifies these constructs. Artificial intelligence monitoring is a structural condition that expands the boundary of legitimate organisational access to information and behaviour. Perceived privacy intrusion (PPI) is the cognitive appraisal of that expansion. Psychological reactance is the motivational response to the autonomy threat the appraisal carries. Turnover intention (TI) is the behavioural form that autonomy restoration takes when restoration inside the organisation appears blocked. These four elements form one causal sequence – condition, appraisal, motivation, behaviour – that serves as the conceptual contribution.
Resolving the question requires conceptual precision about the focal construct. We define it as perceived automated monitoring intensity: The extent to which employees perceive that automated systems track, evaluate and retain their work behaviour. This construct admits a continuum. Key-performance-indicator dashboards anchor its low-opacity, low-inference pole; AI-driven analytics anchor its high-opacity, high-inference pole. We treat AI monitoring as the upper region of this continuum rather than as a categorically separate object, because employees experience the underlying dimension – the perceived scope and unpredictability of automated evaluation – rather than the engineering label.
The study makes three delimited contributions. Firstly, it extends EPM research by specifying a mediated route, replacing direct-effect estimation with an appraisal-then-motivation sequence that names the intervening psychology (Glavin et al., 2024; Ravid et al., 2023). Secondly, it advances algorithmic management scholarship by identifying privacy intrusion as the appraisal that converts opaque governance into resistance, rather than treating resistance as an unexplained reaction to technology (Köchling et al., 2025; Wiener et al., 2023). Thirdly, it extends exit theory by reframing TI as autonomy restoration under constrained voice, which links classic exit reasoning to digitally governed work (Hirschman, 1970; Korunka et al., 2008).
Theoretical review
Turnover intention reflects more than dissatisfaction or workload; it can express a judgement about the legitimacy of organisational control. The structural logic of control begins with disciplinary power. Foucault (1977) argued that visibility produces self-regulation, because individuals who assume continuous observation adjust their own behaviour. Artificial intelligence monitoring intensifies this logic by making observation datafied and continuous rather than spatial and episodic (Leicht-Deobald et al., 2019; Mettler, 2023). We supply the missing bridge in three steps. Observation produces self-regulation; self-regulation experienced as imposed rather than chosen threatens autonomy; and a sustained autonomy threat motivates withdrawal. Disciplinary power therefore explains the structural condition, but the route to turnover runs through the employee’s appraisal of that condition.
Autonomy is central to this route, and self-determination theory justifies its centrality rather than asserting it. Deci and Ryan (2000) demonstrate that autonomy is a basic psychological need whose frustration predicts disengagement and withdrawal. Artificial intelligence monitoring frustrates autonomy in a specific way that distinguishes it from strict EPM. Under observable EPM, an employee can infer the standard and adapt to it. Under opaque AI monitoring, the standard is unknowable, so adaptation becomes impossible, and control becomes unattainable. The threat is not merely ‘being watched’ but ‘being judged by a standard one cannot see’. This unpredictability converts monitoring into a sustained autonomy frustration (Gagné et al., 2022; McAnally & Hagger, 2024).
Privacy theory contributes something autonomy theory cannot. Autonomy concerns control over one’s actions; privacy concerns control over information about oneself. Artificial intelligence’s inferential capacity attacks the second specifically, because it produces information that the employee never disclosed and may not recognise as theirs. Nissenbaum (2004) names this violation through contextual integrity: A privacy harm occurs when information flows exceed the norms appropriate to a context. Artificial intelligence monitoring routinely exceeds those norms by inferring states from behavioural traces. Privacy theory therefore identifies a distinct injury – boundary overreach in the informational domain – that the autonomy concept does not contain (Carter, 2024; Mettler, 2023).
The integration rests on one ordering principle: Cognitive appraisal precedes motivation, and motivation precedes behaviour. Artificial intelligence monitoring establishes the structural condition. Boundary appraisal registers the condition as intrusion when informational norms are violated. Reactance converts intrusion into a motivational state directed at restoring threatened freedom (Brehm, 1966). The AI-specific features sharpen this sequence through a reasoned pairing rather than an enumeration. Opacity and inference act together: Opacity hides the standard while inference widens it, so employees face evaluation that is both expanding and unreadable. Automation and permanence also act together: Automation removes the human discretion that once allowed appeal, while permanence removes the temporal limit that once let evaluations expire. The first pair drives the intrusion appraisal; the second pair removes the conditions under which an employee might restore control internally. Together they convert monitoring from bounded evaluation into unbounded governance (Boyd & Andalibi, 2023; Mettler, 2023).
Hypothesis development
Artificial intelligence monitoring should raise PPI for reasons specific to AI rather than to monitoring in general. Observable EPM collects what employees know they are producing. Artificial intelligence monitoring infers what they have not disclosed and conceals the basis of that inference. Employees therefore appraise the collection as exceeding both what they consented to and what the task requires, which aligns with contemporary definitions of perceived information privacy violation and workplace privacy invasion (Bhave et al., 2020; Li et al., 2025). The opacity of the system compounds the appraisal, because employees cannot verify the limits of access and must assume the broadest plausible scope. Intrusion is thus a predictable response to inference under opacity, not to surveillance as such (Mettler, 2023):
H1: Artificial intelligence monitoring is positively associated with employees’ perceived privacy intrusion.
Artificial intelligence monitoring should also raise psychological reactance through a route that is distinct from the privacy route, which resolves the apparent redundancy between a direct and a mediated path. The privacy route operates through informational appraisal; the direct route operates through behavioural standardisation. When automated systems narrow the range of acceptable behaviour and tighten evaluative feedback, they restrict the employee’s perceived freedom to choose how to work, independent of any informational concern (Feng et al., 2019; Sessions et al., 2023). The two routes are conceptually separable: One threatens control over information, the other threatens control over action. We retain the direct path because the standardisation mechanism does not require the privacy appraisal to operate:
H2: Artificial intelligence monitoring is positively associated with employees’ psychological reactance.
Perceived privacy intrusion should activate reactance because boundary violation is a freedom threat. Privacy-boundary perspectives and communication privacy management theory hold that violations of informational norms provoke defensive arousal and a sense of turbulence when rules about access and use are exceeded (Meng, 2024; Petronio, 2015). When employees judge that monitoring exceeds legitimate authority, they experience the violation as a constraint on personal sovereignty, and reactance follows as the motivational defence of the threatened boundary (Yost et al., 2019). Intrusion is therefore the cognitive trigger that converts exposure into resistance:
H3: Perceived privacy intrusion is positively associated with psychological reactance.
Perceived privacy intrusion should also relate directly to TI, although weaklier, through the erosion of trust. Invasive digital surveillance and privacy violations are linked to reduced organisational trust, higher stress and lower satisfaction, which in turn increase withdrawal and intentions to leave (Kalischko & Riedl, 2024; Li et al., 2025). When monitoring is unavoidable and tied to livelihood, the privacy trade-off appears coercive rather than voluntary, so intrusion influences strategic decisions about leaving rather than discomfort alone:
H4: Perceived privacy intrusion is positively associated with employees’ turnover intention.
Psychological reactance should exert the most proximal influence on TI, and reactance theory must be extended to specify why. Reactance theory describes the motivational state but stops short of naming the behaviour it produces in employment settings. We extend it by specifying the behaviour: When employees perceive no internal channel through which to restore autonomy, exit becomes the available restoration strategy. Sustained reactance under opaque governance therefore converts into withdrawal cognition, because the organisation offers no other route to recover the threatened freedom (Feng et al., 2019; Sessions et al., 2023). This specification is the theoretical advance the hypothesis requires:
H5: Psychological reactance is positively associated with employees’ turnover intention.
The model’s central claim is a serial cascade, and we foreground it rather than testing every possible path. The ordering follows appraisal logic: Cognition precedes motivation, and motivation precedes behaviour. Artificial intelligence monitoring expands informational and evaluative reach; this expansion is appraised as intrusion; intrusion activates reactance; reactance drives the intention to leave. Evidence from related domains shows that privacy-related appraisals and trust judgements often precede and channel effects on downstream outcomes such as commitment, satisfaction and TI, and that sequential mediation models fit such processes better than parallel ones (Kalischko & Riedl, 2024; Zhao et al., 2022). We treat the serial path as the primary explanation and the direct effects as residual mechanisms that the cascade does not fully absorb. This priority answers why serial mediation is theoretically superior to parallel mediation: Parallel mediation treats intrusion and reactance as interchangeable routes, whereas the serial form respects their causal order, in which an appraisal must occur before the motivation it provokes. Specifying this order is what allows the model to explain the process rather than merely to estimate every coefficient:
H6: Perceived privacy intrusion mediates the relationship between AI monitoring and psychological reactance.
H7: Psychological reactance mediates the relationship between AI monitoring and turnover intention.
H8: Perceived privacy intrusion mediates the relationship between AI monitoring and turnover intention.
H9: Psychological reactance mediates the relationship between perceived privacy intrusion and turnover intention.
H10: Artificial intelligence monitoring is indirectly associated with turnover intention through the sequential mediation of perceived privacy intrusion and psychological reactance.
Research design
Sample and procedure
We employed a cross-sectional survey design. The target population comprised Generation Z employees in Indonesian organisations whose work is subject to automated monitoring. We define automated monitoring as the use of software systems that track work activity, evaluate performance from that activity, and retain the resulting data without continuous human mediation. This definition includes AI-driven analytics at the high-opacity pole and key-performance-indicator dashboards at the low-opacity pole, and it excludes purely manual supervision. We adopted this operational boundary to remove the conflation of AI monitoring with non-automated oversight that an undefined sample would permit.
We recruited participants through purposive sampling with a screening filter. Eligibility required three conditions: Birth within the Generation Z range, current employment in an Indonesian organisation and direct exposure to an automated monitoring system as defined above. A screening question at the start of the questionnaire excluded respondents who did not meet all three conditions. We distributed the questionnaire online, informed participants that responses were anonymous and used only for research and stated that no answer was correct or incorrect. We retained 390 usable responses after removing incomplete cases, patterned responding and responses that failed attention checks (see Table 1 for the demographic distribution).
Measures
All constructs were measured using established reflective scales and rated on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The original English items were translated into Indonesian and adjusted to fit the study context without changing their core conceptual meaning. The adaptation involved three steps. Firstly, the wording was simplified so that respondents could understand each statement easily. Secondly, context-specific terms were adjusted to match the research setting and the language normally used by Indonesian respondents. Thirdly, the translated items were reviewed to ensure that each item still represented the same construct as in the original scale. No items were deleted or added during this process. The final Indonesian item wording is reported in Appendix 1 Table 1-A1.
We measured automated monitoring intensity (AMI) with six items grounded in the EPM literature (Stanton, 2000; Stanton & Julian, 2002; Stanton & Weiss, 2000). The items capture the perceived scope, frequency and evaluative use of automated tracking, such as ‘Automated systems are used to monitor my work activities’ and ‘Monitoring data are used in evaluating my performance’. We treat these items as indicators of perceived automated monitoring intensity rather than of verified AI features. This reading is deliberate: Employees respond to the perceived footprint of automated oversight, which is the theoretically relevant exposure. We record the corresponding boundary in the Limitations section, where we note that the items do not certify the presence of specific AI capabilities.
The PPI is measured with three items derived from established privacy scales (Malhotra et al., 2004; Smith et al., 1996) and reworded towards perceived overreach rather than general concern. The adapted items reference the judgement that monitoring exceeds what the work requires, such as ‘Monitoring practices go beyond what is necessary for work purposes’ and ‘I feel uneasy about the way monitoring data are stored or accessed’. The rewording shifts the target from diffuse concern about organisational practices to a contextual appraisal of boundary violation, which is the construct the model requires. We acknowledge the source scales’ origin in privacy concern and treat the distinction as a measurement boundary discussed in the Limitations.
The PRS is comprised of four items adapted from the Hong reactance scale (Hong & Faedda, 1996) and from state operationalisation of reactance (Dillard & Shen, 2005). We anchored each item to the monitoring context, which converts trait stems into situational appraisals, as in ‘Monitoring makes me feel my freedom to choose how I work is restricted’. These anchoring targets state reactance triggered by the present monitoring condition rather than a dispositional tendency. We note in the Limitations that the items draw on a trait instrument and that the contextual anchoring, while standard, does not fully resolve the trait–state distinction.
The TI comes with five items from the validated Turnover Intention Scale (Bothma & Roodt, 2013), which capture cognitive withdrawal, search intention and anticipated exit, such as ‘I often think about leaving this organisation’. The scale records the proximal cognitive state preceding actual turnover behaviour.
Ethical considerations
Ethical clearance to conduct this study was obtained from the Alauddin State Islamic University Research Ethics Code Enforcement Commission (Ref. No. 112/EC/KEPK-UINAM/2025).
Results
We assessed the reflective measurement model before evaluating the structural relationships, following established PLS-SEM guidelines (Hair et al., 2014). VIF values remained below conservative thresholds, which indicates that indicator collinearity did not threaten estimation. Most loadings exceeded 0.70, which supports item reliability. One AI Monitoring item (AMI6) loaded lower at 0.564; we retained it on theoretical grounds because it represents the breadth of the monitored domain. Furthermore, Hair et al. (2014, pp. 113–114) do not advocate the deletion of items if their removal did not materially change composite reliability (CR) or AVE. This remark provides the confidence to retain the item. Table 2 reports the measurement results.
Reliability was acceptable for three constructs. Cronbach’s alpha, rho_A and composite reliability met recommended benchmarks for AI Monitoring, Psychological Reactance and TI, and AVE exceeded 0.50 for every construct, which supports convergent validity. Perceived Privacy Intrusion showed modest internal consistency (alpha = 0.625), above the exploratory threshold but below the conventional 0.70. We report this value transparently and return to it in the Limitations, where we connect it to the construct’s derivation from privacy-concern scales. The measurement model is therefore adequate for structural estimation, with the noted caution for PPI.
We then examined discriminant validity using both the HTMT ratio and the Fornell–Larcker criterion, which together provide a stringent test of construct distinctiveness (Hair et al., 2014). Table 3 reports both.
| TABLE 3: Discriminant validity results (heterotrait–monotrait ratio and Fornell–Larcker). |
Heterotrait–monotrait ratios remained below conservative thresholds, and each construct’s square-root AVE exceeded its correlations with the others, which confirms discriminant validity. The constructs are therefore related but distinct. Global fit was acceptable, with standardised root-mean-square residual (SRMR) values of 0.075 (saturated) and 0.082 (estimated) and normed fit index (NFI) values of 0.780 and 0.769; the Chi-square values (594.760; 622.004) reflect sensitivity to sample size. We then estimated the structural model with 5000 bootstrap resamples, examining the direct and serial indirect paths. Table 4 reports the results, and Figure 1 displays the model.
| TABLE 4: Structural path estimates, indirect effects and explained variance. |
The results support all 10 hypotheses and confirm the cascade’s central ordering. Artificial intelligence monitoring related most strongly to privacy intrusion (β = 0.552), which establishes intrusion as the immediate appraisal of automated oversight. Reactance was the strongest predictor of TI (β = 0.526), and its effect exceeded the direct effect of intrusion on turnover (β = 0.252), which identifies reactance as the proximal driver of exit. The serial indirect path from monitoring through intrusion and reactance to turnover was significant (β = 0.126), and the model explained substantial variance in reactance (R2 = 0.522) and TI (R2 = 0.510). The pattern matches the theoretical priority we assigned to the serial cascade over the parallel routes.
Discussion
Our empirical analysis provides a collective validation of the proposed theoretical framework, as first observed in Hypothesis 1, which posited a positive relationship between AI monitoring and PPI. The findings suggest that the granular, omnipresent nature of AI-driven surveillance transcends traditional electronic performance monitoring, penetrating the private ‘backstage’ of employee behaviour (Mettler, 2023). Theoretically, this aligns with Communication Privacy Management theory, in which organisational technologies can disrupt boundary coordination by automating the collection of information employees perceive as proprietary. Prior work on workplace privacy and people analytics similarly shows that extensive data harvesting is often experienced as intrusive regardless of managerial justification or stated intent (Carter, 2024; Ravid et al., 2020). For practitioners, the implication is that the deployment of AI oversight must be accompanied by rigorous privacy-by-design and ethics-by-design protocols, as the technical capacity to monitor does not equate to moral, legal, or psychological permission to do so (Carter, 2024).
The acceptance of Hypothesis 2 further elucidates the psychological friction generated by AI monitoring, establishing it as a primary antecedent of perceived reactance (PRS). Empirical studies on electronic monitoring and algorithmic management show that workers tend to interpret intensive digital control as a threat to autonomy rather than as a neutral efficiency tool (Jarrahi et al., 2021; Ravid et al., 2023). Within the framework of Psychological Reactance Theory, highly controlling technologies that restrict perceived choice and self-determination reliably elicit reactance responses, and research on surveillance and technostress indicates similar patterns in organisational settings (Yost et al., 2019). The results thus contribute to the burgeoning literature on algorithmic management by showing that loss of agency is a direct byproduct of being governed by opaque systems (Jarrahi et al., 2021). Rather than reducing perceptions of bias, the perceived lack of human empathy and discretion in AI-mediated monitoring can intensify feelings of being ‘trapped’ or unfairly constrained (Mettler, 2023). Practically, organisations may need to design autonomy-supportive AI features, such as explainability, opportunities for input and limited override mechanisms, to mitigate defiant responses and preserve a sense of professional discretion (Parent-Rocheleau & Parker, 2022).
Regarding Hypothesis 3, the significant relationship between PPI and PRS suggests that violations of personal boundaries serve as a cognitive trigger for resistance (Yost et al., 2019). Prior work connects privacy harms to diminished perceived freedom and heightened defensive responses, indicating that privacy is a precondition for psychological freedom rather than a purely passive state (Carter, 2024). Debates over whether reactance is more trait-like or situational increasingly recognise that intrusive surveillance and datafication can generate situational reactance even among generally compliant employees (Yost et al., 2019). Empirical studies on workplace privacy, monitoring, and control suggest that when individuals feel exposed or ‘over-known’, they seek to reassert control by questioning the legitimacy of organisational authority or by covertly undermining imposed constraints (Mettler, 2023; Ravid et al., 2020). For HR and people management, privacy thus functions as an early-warning indicator: Once employees perceive that personal boundaries have been crossed, downstream resistance to performance-enhancing initiatives becomes highly likely, complicating change management and digital transformation efforts (Cardon et al., 2023).
Validation of Hypothesis 4 establishes a direct, albeit moderate, path between PPI and TI. Research on employee surveillance and psychological contracts shows that invasive data practices are often interpreted as a breach of the implicit agreement between employer and employee, undermining trust and commitment and increasing intentions to quit (Miao et al., 2025; Ravid et al., 2023). The classic ‘exit, voice, and loyalty’ logic implies that when employees experience dissatisfaction and feel they lack influence, exit becomes a salient response; empirical work on monitoring and privacy supports the relevance of this model in digitally intensive workplaces (Mettler, 2023). While much consumer research documents a ‘privacy paradox’, studies in employment contexts indicate that when monitoring is unavoidable and linked to livelihood, privacy trade-offs are seen as coercive rather than voluntary (Carter, 2024). Managers therefore need to recognise privacy as a non-monetary job attribute whose erosion can have effects on retention comparable to reductions in pay or benefits (Barthauer et al., 2020; Chen et al., 2024).
Hypothesis 5 yielded one of the strongest effects in the model, indicating that PRS is a potent driver of TI. Work grounded in conservation of resources perspectives suggests that employees leave environments that chronically deplete psychological resources such as autonomy, control and dignity (Barthauer et al., 2020; Chen et al., 2024). Studies of electronic performance monitoring, algorithmic management and technostress consistently link high perceived control loss and surveillance pressure to strain, burnout and withdrawal cognitions (Miao et al., 2025; Ravid et al., 2023). This implies that reactance may be a more proximal predictor of turnover than privacy intrusion itself (Barthauer et al., 2020). This dynamic aligns with recent discussions of increased attrition in highly surveilled, platform-mediated labour markets, where workers often respond to perceived overcontrol by psychologically or physically exiting rather than engaging (Jianu et al., 2025; Zhang et al., 2025). Consequently, retention strategies in AI-mediated workplaces should attend not only to compensation, but also to restoring a sense of voice, discretion and respect in digitally structured workflows (Parent-Rocheleau & Parker, 2022).
Finally, the pattern of indirect and serial mediation reveals a cascading psychological process through which AI monitoring influences TIs. The significant serial pathway from AI monitoring through privacy intrusion and reactance to TI suggests that digital oversight does not simply have a direct effect on exit-related outcomes; instead, it first disrupts privacy boundaries, which then precipitate reactance that ultimately fuels the desire to leave (Ravid et al., 2023; Yost et al., 2019). Similar multistage mechanisms have been observed in research on surveillance, trust and withdrawal, where perceived invasiveness undermines fairness and autonomy beliefs, leading to disengagement and turnover cognitions (Miao et al., 2025). This evidence indicates that interventions focused solely on technical privacy controls, without addressing broader power asymmetries and opacity in algorithmic management, are unlikely to be sufficient (Diro et al., 2025; Heidemann et al., 2024). More broadly, the findings align with calls to view AI implementation as a psychological and ethical intervention, one that reconfigures relationships of control and accountability, rather than as a purely technical upgrade (Mettler, 2023; Zhang et al., 2025). Organisations that neglect these serial effects on privacy, reactance and the psychological contract risk undermining the anticipated productivity benefits of AI through heightened resistance and the loss of critical talent (Jianu et al., 2025; Ravid et al., 2023).
Conclusion
Limitations
This study shows that AI monitoring influences TI indirectly, through PPI and psychological reactance, with reactance as the proximal driver. Artificial intelligence monitoring therefore operates as a governance mechanism whose perceived legitimacy determines whether oversight stabilises performance or erodes retention. The contribution is the specified ordering of appraisal, motivation, and behaviour, which converts an unresolved debate about monitoring’s effects into a testable account of its mechanism.
Several limitations bound these conclusions and follow directly from the measurement choices. Firstly, the AI monitoring items derive from EPM instruments and capture perceived automated monitoring intensity rather than verified AI capabilities; future work should develop items that index opacity, inference, automation and permanence directly. Secondly, the Perceived Privacy Intrusion scale derives from privacy-concern measures and showed modest internal consistency, so its results warrant replication with a purpose-built intrusion scale. Thirdly, the reactance items adapt a trait instrument to a state appraisal through contextual anchoring, which does not fully resolve the trait–state distinction. Fourthly, the cross-sectional design limits causal inference, and self-report may inflate associations. Fifthly, the sample is restricted to Generation Z employees in Indonesia, and industry and contract heterogeneity were not modelled, which limits generalisability.
Future research should address these limits through longitudinal or multi-source designs, behavioural indicators such as system logs and actual turnover and AI-specific measurement. It should also test moderators that the legitimacy account predicts, including transparency, algorithmic fairness, participative implementation and autonomy-supportive leadership. From a managerial perspective, organisations should govern AI monitoring as a socio-technical arrangement, emphasising data proportionality, transparency, employee voice and autonomy-supportive practice, so that oversight does not convert into the reactance that drives talented employees to leave.
Acknowledgements
During the preparation of this work, the authors used ChatGPT (OpenAI) to assist with language refinement and editorial structuring of the article. The content was reviewed and edited by the authors, who take full responsibility for its accuracy.
Competing interests
The author declares that no financial or personal relationships inappropriately influenced the writing of this article.
CRediT authorship contribution
Andi Adawiah: Conceptualisation, Data curation, Formal analysis, Resources, Supervision, Visualisation, Writing – original draft. Nurmal Idrus: Methodology, Project administration, Software, Validation, Writing – original draft, Writing – review & editing. All authors reviewed the article, contributed to the discussion of results, approved the final version for submission and publication and take responsibility for the integrity of its findings.
Funding information
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Data availability
The data supporting the findings of this study are available from the corresponding author, Andi Adawiah, upon reasonable request.
Disclaimer
The views and opinions expressed in this article are those of the authors and are the product of professional research. They do not necessarily reflect the official policy or position of any affiliated institution, funder, agency or that of the publisher. The authors are responsible for this article’s results, findings and content.
References
Bankins, S., Ocampo, A.C., Marrone, M., Restubog, S.L.D., & Woo, S.E. (2024). A multilevel review of artificial intelligence in organizations: Implications for organizational behavior research and practice. Journal of Organizational Behavior, 45(2), 159–182. https://doi.org/10.1002/job.2735
Barthauer, L., Kaucher, P., Spurk, D., & Kauffeld, S. (2020). Burnout and career (un)sustainability: Looking into the blackbox of burnout-triggered career turnover intentions. Journal of Vocational Behavior, 117, 103334. https://doi.org/10.1016/j.jvb.2019.103334
Bhave, D.P., Teo, L.H., & Dalal, R.S. (2020). Privacy at work: A review and a research agenda for a contested terrain. Journal of Management, 46(1), 127–164. https://doi.org/10.1177/0149206319878254
Bothma, C.F.C., & Roodt, G. (2013). The validation of the turnover intention scale. SA Journal of Human Resource Management, 11(1), a507. https://doi.org/10.4102/sajhrm.v11i1.507
Boyd, K.L., & Andalibi, N. (2023). Automated emotion recognition in the workplace: How proposed technologies reveal potential futures of work. Proceedings of the ACM on Human-Computer Interaction, 7(CSCW1), 95. https://doi.org/10.1145/3579528
Brehm, J.W. (1966). A theory of psychological reactance. Academic Press.
Carayon, P. (1994). Effects of electronic performance monitoring on job design and worker stress: Results of two studies. International Journal of Human–Computer Interaction, 6(2), 177–190. https://doi.org/10.1080/10447319409526089
Cardon, P.W., Ma, H., & Fleischmann, C. (2023). Recorded business meetings and AI algorithmic tools: Negotiating privacy concerns, psychological safety, and control. International Journal of Business Communication, 60(4), 1095–1122. https://doi.org/10.1177/23294884211037009
Carter, C. (2024). AI surveillance: Reclaiming privacy through informational control. European Labour Law Journal, 16(2), 245–258. https://doi.org/10.1177/20319525241306327
Chen, G., Wang, J., Huang, Q., Sang, L., Yan, J., Chen, R., Cheng, J., Wang, L., Zhang, D., & Ding, H. (2024). Social support, psychological capital, multidimensional job burnout, and turnover intention of primary medical staff: A path analysis drawing on conservation of resources theory. Human Resources for Health, 22(1), 42. https://doi.org/10.1186/s12960-024-00915-y
Deci, E.L., & Ryan, R.M. (2000). The ‘what’ and ‘why’ of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01
Dillard, J.P., & Shen, L. (2005). On the nature of reactance and its role in persuasive health communication. Communication Monographs, 72(2), 144–168. https://doi.org/10.1080/03637750500111815
Diro, A., Kaisar, S., Saini, A., Fatima, S., Hiep, P.C., & Erba, F. (2025). Workplace security and privacy implications in the GenAI age: A survey. Journal of Information Security and Applications, 89, 103960. https://doi.org/10.1016/j.jisa.2024.103960
Feng, W., Tu, R., Lu, T., & Zhou, Z. (2019). Understanding forced adoption of self-service technology: The impacts of users’ psychological reactance. Behaviour & Information Technology, 38(8), 820–832. https://doi.org/10.1080/0144929X.2018.1557745
Foucault, M. (1977). Discipline and punish: The birth of the prison. Pantheon Books.
Gagné, M., Parker, S.K., Griffin, M.A., Dunlop, P.D., Knight, C., Klonek, F.E., & Parent-Rocheleau, X. (2022). Understanding and shaping the future of work with self-determination theory. Nature Reviews Psychology, 1(7), 378–392. https://doi.org/10.1038/s44159-022-00056-w
Glavin, P., Bierman, A., & Schieman, S. (2024). Private eyes, they see your every move: Workplace surveillance and worker well-being. Social Currents, 11(4), 327–345. https://doi.org/10.1177/23294965241228874
Hair, J.F., Hult, G.T.M., Ringle, C.M., & Sarstedt, M. (2014). A primer on partial least squares structural equation modeling (PLS-SEM). Sage.
Heidemann, A., Hülter, S.M., & Tekieli, M. (2024). Machine learning with real-world HR data: Mitigating the trade-off between predictive performance and transparency. International Journal of Human Resource Management, 35(14), 2343–2366. https://doi.org/10.1080/09585192.2024.2335515
Hirschman, A.O. (1970). Exit, voice, and loyalty: Responses to decline in firms, organizations, and states. Harvard University Press.
Hong, S.-M., & Faedda, S. (1996). Refinement of the Hong Psychological Reactance Scale. Educational and Psychological Measurement, 56(1), 173–182. https://doi.org/10.1177/0013164496056001014
Jarrahi, M.H., Newlands, G., Lee, M.K., Wolf, C.T., Kinder, E., & Sutherland, W. (2021). Algorithmic management in a work context. Big Data & Society, 8(2), 20539517211020332. https://doi.org/10.1177/20539517211020332
Jianu, B., Tussyadiah, I., & Miller, G. (2025). The toll of resisting algorithmic management. International Journal of Contemporary Hospitality Management, 37(12), 4065–4083. https://doi.org/10.1108/IJCHM-12-2024-1821
Kalischko, T., & Riedl, R. (2021). Electronic performance monitoring in the digital workplace: Conceptualization, review of effects and moderators, and future research opportunities. Frontiers in Psychology, 12, 633031. https://doi.org/10.3389/fpsyg.2021.633031
Kalischko, T., & Riedl, R. (2024). On the consequences of electronic performance monitoring in organizations: Theory and evidence. Digital Transformation and Society, 3(1), 50–79. https://doi.org/10.1108/DTS-10-2022-0054
Köchling, A., Wehner, M.C., & Ruhle, S.A. (2025). This (AI)n’t fair? Employee reactions to artificial intelligence (AI) in career development systems. Review of Managerial Science, 19(4), 1195–1228. https://doi.org/10.1007/s11846-024-00789-3
Korunka, C., Hoonakker, P., & Carayon, P. (2008). Quality of working life and turnover intention in information technology work. Human Factors and Ergonomics in Manufacturing & Service Industries, 18(4), 409–423. https://doi.org/10.1002/hfm.20099
Leicht-Deobald, U., Busch, T., Schank, C., Weibel, A., Schafheitle, S., Wildhaber, I., & Kasper, G. (2019). The challenges of algorithm-based HR decision-making for personal integrity. Journal of Business Ethics, 160(2), 377–392. https://doi.org/10.1007/s10551-019-04204-w
Li, J., Wang, Y., Xiong, W., Jiang, Y., Chen, Y., & Yang, Q. (2025). Work autonomy as a buffer: The impact of perceived information privacy violation on job satisfaction and the mediating role of work stress. WORK: A Journal of Prevention, Assessment & Rehabilitation, 82(2), 367–383. https://doi.org/10.1177/10519815251337922
Malhotra, N.K., Kim, S.S., & Agarwal, J. (2004). Internet users’ information privacy concerns (IUIPC): The construct, the scale, and a causal model. Information Systems Research, 15(4), 336–355. https://doi.org/10.1287/isre.1040.0032
McAnally, K., & Hagger, M.S. (2024). Self-determination theory and workplace outcomes: A conceptual review and future research directions. Behavioral Sciences, 14(6), 428. https://doi.org/10.3390/bs14060428
Meng, X. (2024). Conceptualizing and measuring privacy boundary turbulence in technological contexts: Constructing a measurement scale. Information Processing & Management, 61(3), 103658. https://doi.org/10.1016/j.ipm.2024.103658
Mettler, T. (2023). The connected workplace: Characteristics and social consequences of work surveillance in the age of datification, sensorization, and artificial intelligence. Journal of Information Technology, 39(3), 547–567. https://doi.org/10.1177/02683962231202535
Miao, Q., Yin, H., Schwarz, G., & Hussain, M.A. (2025). Disentangling the impact of perceived electronic performance monitoring on employee burnout in the public sector. Public Management Review, 27(12), 3028–3056. https://doi.org/10.1080/14719037.2024.2396080
Nissenbaum, H. (2004). Privacy as contextual integrity. Washington Law Review, 79(1), 119–158.
Parent-Rocheleau, X., & Parker, S.K. (2022). Algorithms as work designers: How algorithmic management influences the design of jobs. Human Resource Management Review, 32(3), 100838. https://doi.org/10.1016/j.hrmr.2021.100838
Petronio, S. (2015). Communication privacy management theory. In C.R. Berger & M.E. Roloff (Eds.), The international encyclopedia of interpersonal communication (pp. 1–9). Wiley.
Ravid, D.M., Tomczak, D.L., White, J.C., & Behrend, T.S. (2020). EPM 20/20: A review, framework, and research agenda for electronic performance monitoring. Journal of Management, 46(1), 100–126. https://doi.org/10.1177/0149206319869435
Ravid, D.M., White, J.C., Tomczak, D.L., Miles, A.F., & Behrend, T.S. (2023). A meta-analysis of the effects of electronic performance monitoring on work outcomes. Personnel Psychology, 76(1), 5–40. https://doi.org/10.1111/peps.12514
Sessions, H., Baer, M.D., Nahrgang, J.D., & Pychlau, S. (2023). From free pastures to penned in: The within-person effects of psychological reactance on side-hustlers’ hostility and initiative in full-time work. Journal of Applied Psychology, 108(12), 1979–1997. https://doi.org/10.1037/apl0001115
Siegel, R., König, C.J., & Lazar, V. (2022). The impact of electronic monitoring on employees’ job satisfaction, stress, performance, and counterproductive work behavior: A meta-analysis. Computers in Human Behavior Reports, 8, 100227. https://doi.org/10.1016/j.chbr.2022.100227
Smith, H.J., Milberg, S.J., & Burke, S.J. (1996). Information privacy: Measuring individuals’ concerns about organizational practices. MIS Quarterly, 20(2), 167–196. https://doi.org/10.2307/249477
Stanton, J.M. (2000). Reactions to employee performance monitoring: Framework, review, and research directions. Human Performance, 13(1), 85–113. https://doi.org/10.1207/S15327043HUP1301_4
Stanton, J.M., & Julian, A.L. (2002). The impact of electronic monitoring on quality and quantity of performance. Computers in Human Behavior, 18(1), 85–101. https://doi.org/10.1016/S0747-5632(01)00029-2
Stanton, J.M., & Weiss, E.M. (2000). Electronic monitoring in their own words: An exploratory study of employees’ experiences with new types of surveillance. Computers in Human Behavior, 16(4), 423–440. https://doi.org/10.1016/S0747-5632(00)00018-2
Sum, C., Shi, C., & Fox, S.E. (2025). ‘It’s always a losing game’: How workers understand and resist surveillance technologies on the job. Proceedings of the ACM on Human-Computer Interaction, 9(2), 1–32. https://doi.org/10.1145/3710902
Wiener, M., Cram, W.A., & Benlian, A. (2023). Algorithmic control and gig workers: A legitimacy perspective of Uber drivers. European Journal of Information Systems, 32(3), 485–507. https://doi.org/10.1080/0960085X.2021.1977729
Yost, A.B., Behrend, T.S., Howardson, G., Darrow, J.B., & Jensen, J.M. (2019). Reactance to electronic surveillance: A test of antecedents and outcomes. Journal of Business and Psychology, 34(1), 71–86. https://doi.org/10.1007/s10869-018-9532-2
Zhang, N., Sun, X., & Jin, C. (2025). Effect of electronic performance monitoring on employees’ job performance: A social information processing perspective. Behavioral Sciences, 15(3), 256. https://doi.org/10.3390/bs15030256
Zhao, X., Wu, C., Chen, C.-C., & Zhou, Z. (2022). The influence of corporate social responsibility on incumbent employees: A meta-analytic investigation of the mediating and moderating mechanisms. Journal of Management, 48(1), 114–146. https://doi.org/10.1177/0149206320946108
Appendix 1
|
Hide
Show all
Email this article (Login required)
