Technostress in the Era of Generative Artificial Intelligence: Mental Health Implications in Medical Education

Abstract

Objective

Generative artificial intelligence (GenAI) has rapidly expanded across education. While its educational benefits are widely recognized, concerns are emerging regarding its potential psychological impact. This narrative review aimed to synthesize current evidence on GenAI-related technostress and associated mental health risks among medical trainees and educators.

Methods

A narrative review was conducted based on a systematic search of PubMed, supplemented by reference list tracking. Publications from January 2010 to January 2026 were considered, with emphasis on studies published after the release of ChatGPT in 2022.

Results

Emerging evidence suggests that GenAI use introduces increased demands, including cognitive monitoring burden, ambiguity regarding appropriate use, and performance-related pressures. These factors may contribute to technostress, digital fatigue, and emotional strain. Medical trainees appear particularly vulnerable due to high-stakes evaluation, academic integrity concerns, and ongoing professional identity formation, whereas educators more commonly experience role overload, techno-uncertainty, and concerns about professional displacement. Conceptual integration with stress–vulnerability and problematic technology use models suggests potential pathways linking GenAI exposure to anxiety, burnout, sleep disturbance, and dependency-like use patterns.

Conclusion

GenAI is emerging not only as a powerful educational tool but also as a potential source of technology-related psychological strain. Informing by psychiatric perspectives strategies and clear institutional guidance may be essential to maximize the benefits of GenAI while mitigating its unintended mental health risks in medical education settings.

INTRODUCTION

Generative artificial intelligence (GenAI) is a new advanced technology that uses machine learning models that generate humanlike expression. There has been a massive increase in the use of GenAI since the introduction of ChatGPT in 2022. Since the release, GenAI has rapidly spread across educational settings including higher and medical education. Students use AI tools for idea creation, assignments, raising concerns about academic integrity of student work [1,2].

Emerging evidence suggests that the rapid integration of AI into daily life and academic environments may carry mental health implications [3]. Recent studies showed that AI-related technostress, driven by factors such as overload or insecurity can be associated with increase in anxiety and depressive symptoms [4]. Also, there is also a growing concern regarding the compulsive ChatGPT use and the relation to sleep disturbance [5].

Previous studies on AI and mental health have focused on the health benefits of AI-based interventions. Little attention has been given to how GenAI functions as a novel stressor in the everyday living and working environments of universities and medical schools [3,4]. From this perspective, students and educators represent both a mentally vulnerable population and users of rapid AI adoption. This raise concerns that AIrelated technostress may contribute to psychiatric outcomes including anxiety, depressive symptoms, burnout, and problematic patterns of use [3-5]. Moreover, GenAI has increased the challenges through AI-assisted cheating and assessment distrust [2], inducing stress, guilt and ethical conflict.

Therefore, this review aimed to integrate evidence on technostress and emotional distress among students and faculty in the era of GenAI. Specifically, we examined how GenAI contributed to technostress and emotional difficulties among students and educators.

METHODS

This review was primarily based on a systematic literature search conducted in PubMed, supplemented by hand-searching through reference list tracking to identify additional relevant studies not captured in the initial database search. We used combinations of the following key terms: “generative artificial intelligence,” “ChatGPT,” “large language models,” “technostress,” “digital fatigue,” “burnout,” “medical students,” “residents,” and “medical educators.” Eligible studies included original research articles, review papers, and conceptual publications addressing GenAI or AI-related technostress, mental health outcomes, or technology-related psychological strain. Studies that did not report mental health or technostress-related outcomes were excluded. It covered publications from January 2010 to January 2026, reflecting development of technostress research and the emergence of GenAI. Studies published after 2022 were prioritized for the relevance to the post GenAI era.

CONCEPTUAL FRAMEWORK

Technostress

Technostress has been defined as a multidimensional stress framework directly related to information and communication technology (ICT), and the five core factors suggested by Tarafdar [6] is widely used: 1) Techno-overload, 2) Techno-invasion, 3) Techno-complexity, 4) Techno-insecurity, 5) Techno-uncertainty [7]. Techno–overload, invasion, complexity, insecurity, and uncertainty refer to excessive workload, pressure due to constant connectivity, complexity and perceived lack of competence. These components together form the core dimensions of technostress that increase job stress, fatigue, and burnout risk [6,8].

Digital fatigue

Digital fatigue is a broad construct that explains how continued exposure to digital environments contributes to technostress. This is referred to a state in which excessive information overload and persistent connectivity deplete cognitive and emotional resources [4,9].

Overload of information happens when messages arrive simultaneously from channels such as email or apps, leading to exceeding an individual’s processing capacity. Continued connection is related to social and organizational pressures to always remain in touch. These lead to fatigue, tiredness, irritability, shortened sleep duration, and increased job or academic stress [10-12].

AI/GenAI stress

GenAI-specific technostress refers to stressors arising not merely from general ICT use but from sustained interaction with large language models and other generative AI tools in academic and professional settings [4,13]. It includes the burden of identifying and correcting AI outputs such as monitoring hallucinations or inaccuracies [2].

Beyond classic technostress, emerging concepts capture AI-specific anxiety. AI anxiety reflects fear and apprehension related to the spread of AI systems [3,14]. The fear of automation focuses on perceived risk of job displacement and skill devaluation [7].

From a psychiatric perspective, these technology-related stressors may interact with individual vulnerability to increase the risk of anxiety, depressive symptoms, and burnout.

GenAI USE AND MENTAL HEALTH IN MEDICAL TRAINEES

Medical students and residents are increasingly using GenAI tools for exam preparation, drafting or editing academic documents, and combining these tools across both preclinical and clinical training tasks [15,16]. In surveys, trainees relatively perceive GenAI as helpful for improving efficiency, organizing information, and differential diagnosis, while also expressing concerns about hallucinations, bias, and ethical implications of AI use [16,17].

Many students report that expectations from peers and their institution can make usage of GenAI as normative. This may potentially create pressure, and choosing not to use is viewed as inefficient or competitively disadvantageous in exams or clinical tasks [16-18].

Given that direct evidence on problematic GenAI use remains limited, findings from broader digital technology research are informative [5,19]. Previous studies show that excessive or compulsive use of smartphones, social media, and other online academic tools is associated with higher anxiety, sleep disturbance, burnout, and depressive symptoms among university and medical students [5,20,21].

Educational psychology research suggests although cognitive unburdening can ease short-term mental load, heavy or unreflective reliance on GenAI may be associated with lower self-efficacy, reduced self-control over learning and weaker development of cognitive skills [4,22]. These findings suggest that when trainees use GenAI primarily as a replacement rather than a scaffold for their academic work, there may be a risk of increasing concern about inappropriate AI use, and greater vulnerability to anxiety and burnout [5,19].

In the GenAI era, blurred boundaries between authentic AI-assisted learning and academic misconduct may induce trainees to feel uncertain about acceptable use, potentially contributing to chronic stress and internalized shame [18,23].

These issues may be prominent for medical students and residents. They learn at clinical environments that require diagnostic accuracy, patient safety, and professional identity formation [24,25]. Regarding this context, uncertainty regarding appropriate GenAI use may carry further cognitive and ethical issues compared to general higher education settings [26]. These patterns may represent emerging forms of technology-related psychological strain in medical trainees.

GenAI USE AND MENTAL HEALTH IN MEDICAL EDUCATORS

Medical educators and clinical professors are easily confronted to work overload because they are frequently required to educate, see patients and do research at the same time [27]. This kind of complex job requirements need longer hours leading to emotional burnout and decrease in self-efficacy [28]. In a study of pediatric professors, it was found that although the relationship with students were fruitful, balancing between education and clinical work, insufficiency of job hours were the reasons of burnout [29].

Studies on higher education found that the spread of digitalization, online classes, real time communication tools require educators to adapt to the technology promptly [7]. This may increase the risk of technostress such as techno-overload or techno-uncertainty.

After the introduction of GenAI, educators experience additional AI-specific burden over the existing technostress [30]. First, the pressure to detect AI-assisted cheating and plagiarism in assessments and reports may work as a burden [1,2]. Second, expectations to redesign assessment around GenAI combined with limited time and institutional support burdens on educators who are already overloaded [20]. Third, some educators who feel less proficient with GenAI than their students or younger colleagues may experience a competence gap. This may have negative effects on self-efficacy and increase anxiety about their future job security as educators, as well as a sense of threat in being displaced from the role of supervisors [31,32].

Earlier studies in medical and health professions education show that medical educators primarily used GenAI as just a support tool for lecture preparation, drafting exam items, and composing feedback comments in terms of time saving [33]. At the same time, many educators express concerns about the accuracy of AI outputs, limitation of critical reasoning, and the potential misuse of students, and the risk of undermining professional expertise [34]. These findings show the emotional burden that educators experience as they try to balance assessment fairness with their professional identity, suggesting that GenAI may be added as an additional source of burnout and technostress among medical educators [32].

From a psychiatric perspective, these accumulating demands may contribute to chronic occupational stress and heighten vulnerability to burnout, anxiety, and emotional exhaustion among medical educators.

INTEGRATIVE PSYCHIATRIC FRAMEWORK

The proposed integrative framework conceptualizes GenAI–related mental health risks along a pathway from demands and stressors to technostress, then to burnout, and ultimately to diseases such as anxiety disorder or depression.

Burnout and mental health outcomes in medical trainees and physicians

Meta-analyses and systematic reviews in physicians and medical students link burnout and higher levels of depressive and anxiety symptoms, sleep problems [27,28]. Theses may support a pathway from chronic occupational or academic stress through burnout to more formal psychiatric outcomes. These findings suggest that GenAI-related technostress may add onto existing educational and clinical demands, which may exacerbate the risk of already vulnerable groups such as medical students [7].

Stress, control, and coping perspectives

Classic stress and coping theories suggest that individuals experience greater psychological harm when high demands and low control coexist, and when there are inadequate coping resources [35]. In the era of GenAI, educators and trainees must navigate rapid technological changes, face fears about competence or replacement, which can increase the perceived loss of control. This may intensify technostress and burnout, particularly when there is limitation of institutional guidance or psychosocial support [5,7].

Problematic use and behavioral addiction perspectives

Previous research on problematic internet use and behavioral addictions provides a useful complementary framework for interpreting compulsive or dependency-like patterns of GenAI use, especially when it is employed as a coping strategy under high stress [36]. Model concepts previously presented by previous research showed cycles of short-term relief through excessive use, loss of control, and subsequent guilt [37]. This concept may be related to emerging reports of compulsive ChatGPT use associated with anxiety, burnout, and sleep disturbance in students and professionals [5]. Together, these ideas support viewing GenAI not only as a supporting tool, but also as a potential stressor and object of problematic use, emphasizing the need for psychiatric models that integrate demands, technostress, burnout, and addiction-like dynamics.

DISCUSSION

This narrative review integrates emerging evidence suggesting the evolvement of GenAI from a time-reducing educational tool into a notable psychosocial stressor within academic environments. In the reviewed literature, GenAI was related to demands such as monitoring burden, ambiguity regarding legitimate use, that may induce technostress and emotional strain among both medical trainees and educators [2,18,32,34].

These findings can be understood within existing stress-vulnerability frameworks from a psychiatric perspective. Rapid technological changes, unclear norms, may reduce self-control while amplifying stress responses [38]. In parallel, the repetitive use of GenAI can drive digital fatigue and cumulative cognitive load, both of which are known to be related to emotional exhaustion and burnout [39]. Especially, the literature on the problematic use of internet suggests that some individuals may engage in excessive or overload GenAI use that follows patterns resembling dependency-like behaviors [40,41].

This review also tried to differentiate the impact of GenAI between medical trainees and educators. Medical trainees operate within high-pressure academic environments in which maintaining academic integrity and developing professional identity are critical developmental tasks [19]. In contrast, educators are faced with role overload, techno-uncertainty, and concerns regarding professional displacement [7,32]. These stress profiles overlap partially, but are distinct, and need population-sensitive interventional strategies.

The present integration shows several clinical and educational implications. First, technostress and AI-related burden may represent emerging needs for mental health surveillance withing medical education settings. Second, institutions may benefit from developing clear guidelines for appropriate GenAI use. Third, psychiatric wellness programs in academic centers should incorporating digital stress management and coping skills training.

In addition to individual coping strategies, GenAI-related technostress may be mitigated by establishing clear interventions from institutional perspectives. Medical schools and academic hospitals may benefit from developing structured guidelines that address appropriate and ethical GenAI use in learning and assessment contexts. Faculty programs may help educators to get used to rapid development of AI technologies and reduce techno-uncertainty. Furthermore, education initiatives that increase awareness of digital fatigue, support mental health on technology-related stressors, and integrate digital literacy and responsible AI use for students may help develop balanced approaches to GenAI adaptation. These strategies may help learners and educators to perceive GenAI as supportive resource while protecting mental well-being.

This review has several limitations. As a narrative review, the synthesis is restricted due to heterogeneity in study design and much of the current evidence remains cross-sectional. Longitudinal data on problematic GenAI use and significant psychiatric outcomes are still limited. In addition, the rapid development of GenAI technologies means that the risk is likely to shift continuously. Future research should be done in longitudinal designs and use AI-specific technostress measures.

In conclusion, GenAI is emerging not only as a powerful educational tool, but also as a potential source of technology-related psychological strain. Through pathways like perceived loss of control and cognitive overload, GenAI may contribute to technostress and further mental health risks. Approaches through psychiatric perspectives may help maximize the educational and practical benefits of GenAI while mitigating its unintended psychological impact.