Previous Article in Journal
A Mixed-Methods Assessment of the Maternal Death Surveillance Response in Tanzania as an Information Source Beyond the Numbers: An Opportunity for SDG 2030 Tracking
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Trends in Public Health
Volume 1
Issue 2
10.3390/tph1020015
tph-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Artificial Intelligence Trust as a Buffer Against Stress: Implications for Mental Toughness and Student Well-Being

by
Akanksha Anand
1,*,
Vishnu Kumar
2 and
Krishna Bista
3
1
School of Community Health and Policy, Morgan State University, Baltimore, MD 21251, USA
2
Department of Industrial and Systems Engineering, Morgan State University, Baltimore, MD 21251, USA
3
Department of Advanced Studies, Leadership and Policy, School of Education and Urban Studies, Morgan State University, Baltimore, MD 21251, USA
*
Author to whom correspondence should be addressed.
Trends Public Health 2026, 1(2), 15; https://doi.org/10.3390/tph1020015
Submission received: 2 June 2026 / Revised: 31 July 2026 / Accepted: 26 August 2026 / Published: 10 September 2026
Browse Figure
Versions Notes

Abstract

Background: Stress represents a significant public health concern and is associated with diminished psychological resilience among students and young adults. Mental toughness, the capacity to function effectively under pressure, may be influenced by attitudes toward emerging technologies. Methods: This cross-sectional study investigated whether trust in artificial intelligence (AI) moderates the relationship between perceived stress and mental toughness. Students attending a Historically Black College or University (HBCU) participated (n = 109); 84.4% were aged 18–25 years. Perceived stress was assessed using an adapted six-item DASS stress index, while mental toughness and AI trust were measured using adapted multi-item scales. Hierarchical regression analyses examined the moderating role of AI trust. Results: Higher perceived stress was associated with lower mental toughness. Although AI trust did not directly predict mental toughness after accounting for the interaction, a significant interaction emerged: students with greater AI trust showed a weaker negative association between stress and mental toughness. Conclusions: Higher AI trust was associated with attenuation of the negative relationship between perceived stress and mental toughness among students. These findings may inform future research on trustworthy AI, student well-being, and evidence-based public health strategies for students in high-stress environments.

1. Introduction

Stress is a significant public health concern, particularly among students and young adults who face increasing academic, financial, and social pressures. Elevated stress levels have been consistently associated with adverse mental health outcomes, including anxiety, academic burnout, reduced psychological resilience, and diminished well-being among college students [1,2,3,4,5,6,7]. As mental health challenges continue to increase among college populations, identifying factors that promote resilience has become an important public health priority. Recent evidence suggests that resilience-focused approaches can improve university students’ mental health and well-being [8,9]. Mental toughness reflects an individual’s capacity to sustain performance, maintain confidence, and manage adversity under pressure, making it an important construct for understanding resilience in stressful environments [10,11,12,13].
At the same time, artificial intelligence (AI) technologies are becoming increasingly integrated into educational, occupational, and healthcare environments. AI-enabled systems, including virtual assistants, decision-support tools, and digital mental health platforms, are transforming how individuals access information, make decisions, and manage health-related concerns [14,15,16,17]. Recent evidence suggests that AI-based interventions have the potential to improve access to mental health resources, provide scalable support, and enhance psychological well-being, particularly in settings where traditional services are limited [18,19,20,21].
Despite these potential benefits, public trust in AI remains inconsistent. Concerns regarding algorithmic bias, lack of transparency, privacy risks, and ethical accountability continue to influence perceptions of AI systems and may limit their adoption [22,23,24]. Individuals often demonstrate a pattern of cautious acceptance, simultaneously recognizing the utility of AI while expressing concerns about reliability, fairness, and the broader societal implications of automation [25]. As AI technologies become increasingly involved in health-related decision-making and psychological support, understanding how trust influences engagement with these systems is essential.
Although research examining AI adoption and digital mental health continues to expand, human trust in AI has been recognized as a critical factor influencing the acceptance and effective use of AI technologies across diverse settings [26]. However, limited attention has been given to the role of AI trust in shaping psychological resilience under stressful conditions. Addressing these gaps is important because digital technologies are increasingly positioned as tools for supporting mental health, stress management, and psychological resilience [27]. Students attending Historically Black Colleges and Universities (HBCUs) may experience unique academic, financial, and social stressors while also benefiting from culturally supportive educational environments. Understanding factors that promote resilience and psychological well-being among HBCU students is therefore an important public health priority. Understanding the relationship between stress, resilience, and AI trust may provide valuable insights for developing effective and equitable public health interventions.
Therefore, the purpose of this study is to examine whether trust in AI moderates the relationship between perceived stress and mental toughness among students attending a Historically Black College or University (HBCU). By investigating AI trust as a potential buffering factor, this study contributes to emerging discussions regarding the role of AI in supporting psychological resilience and mental health within public health contexts.

2. Conceptual Background

2.1. Conceptualizing Stress and Mental Toughness

Stress has long been recognized as a major determinant of mental health and well-being. According to stress and coping theory, individuals experiencing high levels of stress are more vulnerable to psychological strain and reduced adaptive functioning [3]. Previous research consistently demonstrates a negative association between stress and mental toughness, indicating that individuals experiencing greater stress often report lower resilience, diminished emotional regulation, and reduced ability to cope effectively with challenges [10,11]. Mental toughness reflects an individual’s capacity to sustain performance, maintain confidence, and manage adversity under pressure, making it an important construct for understanding resilience in stressful environments. Recent evidence among U.S. college students further indicates that mental toughness is positively associated with grit and may be strengthened through engagement in regular moderate-to-vigorous physical activity, highlighting that resilience-related characteristics are modifiable rather than fixed [12]. This perspective is consistent with work conceptualizing mental toughness as a capacity that can be assessed and developed rather than treated solely as fixed [13].

2.2. Trust in Artificial Intelligence

Trust has emerged as a critical factor influencing individuals’ willingness to engage with AI-enabled systems. Trust in AI encompasses perceptions of reliability, fairness, transparency, and dependability, all of which contribute to the acceptance and continued use of technological systems [14]. However, concerns regarding algorithmic bias, explainability, privacy, and ethical governance continue to shape public attitudes toward AI, particularly in sensitive domains such as healthcare and mental health services [22,24]. Recent evidence suggests that trust in AI health applications varies considerably across populations and is strongly influenced by perceptions of transparency, usability, and ethical safeguards [28].
Individuals frequently exhibit ambivalent attitudes toward AI technologies, balancing perceived benefits with concerns regarding potential risks and unintended consequences [25,29]. These dynamics highlight the importance of examining AI not only as a technological innovation but also as a psychological and social factor that may influence behavior, decision-making, and well-being. Importantly, trust may also shape how individuals perceive and engage with AI-enabled resources during periods of stress. When AI systems are perceived as reliable, transparent, and fair, individuals may be more willing to engage with these technologies [14,28]. This suggests that AI trust may have relevance not only for technology acceptance but also for understanding how individuals respond to stress and maintain psychological resilience.

2.3. Technology Acceptance Model and AI Trust

The Technology Acceptance Model (TAM) provides a widely used framework for understanding technology adoption [30,31]. TAM proposes that perceived usefulness and perceived ease of use are primary determinants of an individual’s intention to adopt and use new technologies. While these constructs explain initial acceptance, they do not fully capture individuals’ willingness to rely on technology in situations involving uncertainty, vulnerability, or personal decision-making.
In the context of AI, trust extends beyond traditional acceptance factors by influencing whether users feel comfortable depending on AI-generated information and recommendations. Even when AI systems are perceived as useful, concerns regarding fairness, privacy, transparency, and accuracy may limit engagement. Consequently, trust plays a central role in shaping both the adoption and sustained use of AI-enabled systems.
Extending TAM in the present study, AI trust is conceptualized not only as a factor related to technology acceptance [30,31] but also as a contextual factor that may influence how individuals respond to stress. This distinction provides a basis for examining whether the relationship between perceived stress and mental toughness varies according to an individual’s level of AI trust.

2.4. Proposed Moderating Role of AI Trust

Emerging theoretical perspectives conceptualize AI as an augmentative technology designed to support, rather than replace, human capabilities [29]. In high-stress environments, AI-enabled tools may assist individuals by reducing cognitive burden, providing information and guidance, and supporting adaptive coping strategies. These functions may be particularly relevant within digital mental health interventions, where AI systems are increasingly used to deliver personalized support, symptom monitoring, and behavioral guidance [20,21,32,33,34].
However, the effectiveness of AI-enabled support depends largely on whether individuals trust these systems. Users who perceive AI technologies as reliable and trustworthy may be more likely to engage with available resources, utilize recommendations, and derive psychological benefits from their use. Conversely, individuals with low levels of trust may be less likely to use AI tools effectively, potentially limiting their value as coping resources.
From this perspective, AI trust may function as a contextual factor that shapes the relationship between stress and resilience. Rather than directly increasing mental toughness, trust in AI may influence how strongly stress affects an individual’s ability to cope with challenges. Consequently, individuals with higher levels of AI trust may experience a weaker negative association between stress and mental toughness than those with lower levels of trust.
Taken together, this conceptual framework provides a theoretical rationale for examining AI trust as a moderator, rather than solely as an independent predictor, of the relationship between perceived stress and mental toughness. Specifically, greater trust in AI may facilitate engagement with AI-enabled resources during periods of stress, potentially attenuating the negative association between perceived stress and mental toughness [14,29].
Accordingly, the following hypotheses are proposed:
H1. 
Higher levels of perceived stress will be associated with lower levels of mental toughness.
H2. 
Higher levels of AI trust will be associated with higher levels of mental toughness.
H3. 
AI trust will moderate the relationship between stress and mental toughness such that the negative effect of stress on mental toughness will be weaker among individuals with higher levels of AI trust.

3. Materials and Methods

3.1. Study Design and Participants

A cross-sectional survey study was conducted to examine the relationships among perceived stress, AI trust, and mental toughness among students attending a Historically Black College or University (HBCU). Participants were recruited from a Historically Black College or University (HBCU) using a convenience sampling approach through undergraduate courses and university listservs. Eligible participants were required to be at least 18 years of age and currently enrolled as students at a Historically Black College or University (HBCU). Although recruitment primarily targeted undergraduate courses, a small number of graduate students also participated and were retained in the analyses.
Data were collected through an anonymous online survey. Participation was voluntary, and no personally identifying information was collected. A total of 109 students completed the survey. The use of a cross-sectional survey design is consistent with established public health and behavioral science research examining psychological constructs and their interrelationships among college student populations [1,2].
The study was conducted in accordance with the ethical principles of the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of Morgan State University (IRB Protocol #26/02-0078). Electronic informed consent was obtained from all participants prior to survey completion.

3.2. Measures

3.2.1. Perceived Stress

Perceived stress was assessed using six items adapted from the stress subscale of the Depression Anxiety Stress Scales (DASS-21) [4]. The items assessed difficulty winding down, over-reactivity, nervous energy, agitation, difficulty relaxing, and touchiness/irritability. Participants rated the extent to which each statement applied to them during the previous week using a four-point Likert scale ranging from 0 (“did not apply to me at all”) to 3 (“applied to me very much or most of the time”). Higher scores indicated greater perceived stress. Because the standard DASS-21 stress subscale contains seven items and the present survey included six, the measure is described as an adapted six-item DASS stress index rather than the complete DASS-21 stress subscale. In the present study, the adapted stress index demonstrated good internal consistency (Cronbach’s α = 0.82).

3.2.2. Mental Toughness

Mental toughness was assessed using an adapted six-item measure informed by the mental toughness literature, including the work of Clough et al. [10] and Gucciardi [11]. The items assessed participants’ ability to remain calm under pressure, maintain a sense of control, persist in achieving goals, view challenges as opportunities for growth, believe in their ability to succeed, and navigate social situations effectively. Participants responded using a Likert-type scale, with higher scores indicating greater mental toughness. The adapted six-item measure demonstrated good internal consistency (Cronbach’s α = 0.81).

3.2.3. Perceived AI Trust and Acceptance

Perceived AI trust and acceptance were measured using four items adapted from the Technology Acceptance Model (TAM) [30,31,35] and the human trust in artificial intelligence literature [14]. The retained items assessed participants’ trust in AI decision-making, perceptions of AI fairness, openness to using AI technologies in daily life, and willingness to recommend AI-powered services. Participants responded using a five-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree), with higher scores indicating greater perceived AI trust and acceptance. Based on the exploratory factor analysis, the two negatively worded items assessing concerns about job displacement and discomfort with autonomous AI decision-making did not load adequately on the primary factor and were excluded from the final scale. Consequently, only the four retained items were used to compute the composite score and internal consistency reliability. The final four-item scale demonstrated good internal consistency (Cronbach’s α = 0.81). The full survey instrument is provided in Supplementary File S1.

3.3. Validity and Reliability

The questionnaire items were adapted from established literature on technology acceptance, trust in AI, mental toughness, and stress [4,10,11,12,14,30,31,36]. Content validity was supported through literature-based item adaptation and review for relevance to the study objectives. Construct validity was evaluated using exploratory factor analysis (EFA), with sampling adequacy assessed using the Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test of sphericity. Internal consistency reliability was assessed using Cronbach’s alpha, with values ≥ 0.70 considered acceptable [37].
Exploratory factor analysis supported the one-dimensionality of the perceived AI trust and acceptance construct (KMO = 0.658, Bartlett’s test: χ2(6) = 179.05, p < 0.001). The four retained items loaded on a single factor, with loadings ranging from 0.497 to 0.927 and explaining 54.88% of the total variance. The scale demonstrated good internal consistency (Cronbach’s α = 0.810).

3.4. Data Analysis

Descriptive statistics were first computed to characterize the study sample and summarize participant responses. Frequencies and percentages were calculated for categorical demographic variables, including age group, gender identity, academic level, field of study, and employment status. Means, standard deviations, and score distributions were subsequently calculated for the primary study variables, including perceived stress, AI trust, and mental toughness. These descriptive analyses provided an overview of participant characteristics and the distribution of key constructs prior to hypothesis testing. Cases with missing data on one or more study variables were excluded using listwise deletion. Consequently, the correlation and regression analyses were conducted using an analytic sample of 103 participants.
Sociodemographic variables were not included as covariates because the study hypotheses focused on the relationships among perceived stress, AI trust, and mental toughness, and the modest analytic sample limited the number of parameters that could be estimated reliably without reducing statistical precision. Future studies should examine these relationships after adjustment for relevant demographic characteristics.
Pearson correlation analyses were conducted to examine bivariate relationships among perceived stress, AI trust, and mental toughness. To test the study hypotheses, hierarchical multiple regression analyses were performed. Predictor variables were mean-centered prior to analysis to reduce multicollinearity, and an interaction term (Stress × AI Trust) was created to assess moderation effects. In Model 1, stress and AI trust were entered as main-effect predictors of mental toughness. In Model 2, the interaction term was added to determine whether AI trust moderated the relationship between stress and mental toughness. Statistical significance was evaluated at p < 0.05. All analyses were conducted using IBM SPSS Statistics for Macintosh, Version 31.0 (IBM Corp., Armonk, NY, USA).

4. Results

4.1. Participant Characteristics

A total of 109 students participated in the study (Table 1). The sample was predominantly composed of traditional college-aged students, with 84.4% between 18 and 25 years of age. Female participants represented more than two-thirds of the sample (68.8%), while males accounted for 29.4%; 1.8% preferred not to disclose their gender identity.
Regarding academic standing, the largest proportion of participants were senior undergraduate students (54.1%), followed by junior students (21.1%) and graduate students (14.7%). Most respondents were enrolled in STEM-related disciplines (74.3%). Employment was common among participants, with 58.8% reporting either full-time or part-time employment while attending school and 41.3% reporting no current employment.

4.2. Descriptive Statistics and Correlations

Descriptive statistics and Pearson correlation coefficients for the primary study variables are presented in Table 2. Mental toughness was negatively correlated with perceived stress (r = −0.340, p < 0.001), indicating that higher levels of stress were associated with lower levels of mental toughness. AI trust demonstrated a small positive correlation with mental toughness (r = 0.187, p = 0.048); however, this finding should be interpreted cautiously because it was close to the conventional threshold for statistical significance. No significant relationship was observed between stress and AI trust (r = −0.006, p > 0.05).

4.3. Hierarchical Regression Analysis

A hierarchical multiple regression analysis was conducted to examine whether AI trust moderated the relationship between stress and mental toughness (Table 3). In Model 1, stress and AI trust were entered as predictors of mental toughness. The overall model was statistically significant, F(2, 100) = 8.80, p < 0.001, explaining 15.0% of the variance in mental toughness (R2 = 0.150). Stress was a significant negative predictor of mental toughness (β = −0.339, p < 0.001), supporting Hypothesis 1. AI trust was a significant positive predictor in Model 1 (β = 0.185, p = 0.048), providing limited preliminary support for Hypothesis 2 because the association was modest and close to the conventional threshold for statistical significance.
When the interaction term (Stress × AI Trust) was added in Model 2, the model remained statistically significant, F(3, 99) = 7.59, p < 0.001, accounting for 18.7% of the variance in mental toughness (R2 = 0.187). The addition of the interaction term resulted in a significant increase in explained variance (ΔR2 = 0.037, p = 0.036). The corresponding local effect size was small (Cohen’s f2 = 0.046).

4.4. Moderating Effect of AI Trust

The interaction between stress and AI trust was statistically significant (β = 0.201, p = 0.036). As shown in Figure 1, mental toughness decreased as perceived stress increased; however, the decline was less pronounced among individuals reporting higher levels of AI trust. In contrast, participants with lower levels of AI trust exhibited a steeper decline in mental toughness as stress increased. These findings indicate that AI trust attenuates the negative relationship between stress and mental toughness, suggesting a buffering effect.

5. Discussion

This study examined the relationships among perceived stress, trust in artificial intelligence (AI), and mental toughness among students attending a Historically Black College or University (HBCU). Consistent with Hypothesis 1, perceived stress was negatively associated with mental toughness, indicating that students experiencing higher levels of stress reported lower resilience. Although AI trust demonstrated a modest positive association with mental toughness in the initial model, its direct effect was no longer significant after accounting for the interaction term. Most importantly, AI trust moderated the relationship between stress and mental toughness, such that the negative impact of stress was attenuated among individuals reporting higher levels of AI trust.

5.1. Stress and Mental Toughness

The finding that stress was negatively associated with mental toughness is consistent with previous research demonstrating that elevated stress impairs resilience, emotional regulation, and adaptive functioning [2,3,4]. Lazarus and Folkman’s stress and coping framework suggests that stress occurs when individuals perceive environmental demands as exceeding their available coping resources [3]. Under such conditions, prolonged stress may reduce psychological well-being and compromise the ability to effectively manage challenges. The present findings also align with research identifying mental toughness as a key component of resilience that enables individuals to maintain focus, confidence, and performance under pressure [10,11].
The present findings are also consistent with emerging evidence suggesting that mental toughness is influenced by modifiable psychological and behavioral factors. Stamatis et al. [12] reported a positive association between grit and mental toughness among U.S. students, with this relationship further strengthened among individuals who engaged in at least 75 min of moderate-to-vigorous physical activity per week. Their findings suggest that mental toughness is not solely a fixed personality characteristic but can be enhanced through supportive behaviors and environmental influences. Consistent with this perspective, the current study identifies trust in AI as a contextual factor that may help individuals maintain resilience in the face of elevated stress. Together, these findings highlight multiple pathways through which resilience can be strengthened among college students.
Given the increasing prevalence of stress among college students, strengthening resilience remains an important public health priority [1,2]. Understanding the factors that help individuals maintain mental toughness under stressful conditions may contribute to the development of more effective interventions aimed at improving student well-being and long-term mental health outcomes.

5.2. AI Trust as a Buffer Against Stress

A central contribution of this study is the finding that AI trust moderated the relationship between stress and mental toughness. Although AI trust did not independently predict mental toughness after accounting for the interaction effect, higher levels of trust were associated with a weaker negative relationship between stress and mental toughness. This pattern suggests that AI trust may function as a contextual correlate of resilience under stress; however, longitudinal research is needed to clarify the mechanisms underlying the association.
The results are consistent with prior literature emphasizing the importance of trust in determining individuals’ willingness to engage with AI-enabled systems [14]. Confidence in the reliability and transparency of AI technologies may influence whether individuals are willing to engage with and rely upon these systems [30,31,35,38]. Emerging perspectives further characterize AI as an augmentative technology designed to enhance human capabilities rather than replace them [29]. In this context, trusted AI systems may provide informational, cognitive, or decision-support resources that reduce uncertainty and facilitate adaptive coping responses during periods of stress [15].
These findings are particularly relevant within digital mental health settings, where AI-enabled interventions are increasingly used to provide psychological support, symptom monitoring, and personalized guidance [20,21,39]. Recent studies have highlighted the potential of AI chatbots and digital mental health platforms to improve access to care, increase service availability, and support psychological well-being at scale [21,33,34]. However, the effectiveness of such systems depends heavily on users’ willingness to trust and engage with them [28]. The present findings reinforce the view that trust is not merely a technological consideration but also a psychological factor influencing the benefits individuals derive from AI-supported resources. Similar to the role of physical activity in strengthening the relationship between grit and mental toughness among students [12], trust in AI may function as a psychological and technological resource that attenuates the adverse effects of stress and supports adaptive coping. These findings contribute to the growing literature suggesting that resilience is shaped not only by individual characteristics but also by access to supportive resources and environments.

5.3. Public Health Implications

From a public health perspective, these findings underscore the importance of promoting trustworthy and ethical AI systems. As AI technologies become increasingly integrated into healthcare, education, and mental health services, concerns related to transparency, fairness, privacy, and accountability remain significant barriers to adoption [22,24]. Previous research has shown that individuals may resist AI-based health interventions when they perceive risks related to accuracy, bias, or reduced human involvement in decision-making [32].
The HBCU context is also important when interpreting these findings. Students attending Historically Black Colleges and Universities often navigate multiple stressors that extend beyond traditional academic demands, including financial pressures, employment responsibilities, family obligations, and broader social and structural challenges that may affect mental health and well-being. At the same time, HBCUs have historically served as important environments for academic support, cultural affirmation, and educational opportunity for underserved populations [40]. These unique contextual factors may influence both resilience and students’ engagement with emerging technologies.
In this context, trusted AI systems may serve as supplementary support resources by providing timely access to information, stress-management tools, and mental wellness guidance. The findings suggest that enhancing public trust may increase the effectiveness of AI-enabled interventions aimed at supporting resilience and psychological well-being. Efforts to improve AI literacy, transparency, and ethical governance may facilitate more equitable engagement with AI technologies and maximize their public health benefits [22,24].

5.4. Limitations and Future Research

Several limitations should be acknowledged. First, the cross-sectional design precludes causal inference regarding the relationships among stress, AI trust, and mental toughness. Second, all variables were measured using self-report instruments, which may be subject to response bias. Third, participants were recruited from a single HBCU, which may limit the generalizability of the findings to other student populations and institutional contexts. Fourth, the study measured attitudes toward AI rather than actual use of AI-enabled mental health or support tools; therefore, the observed interaction should not be interpreted as evidence that AI use itself caused the buffering pattern. AI trust may also reflect broader technology attitudes or other unmeasured psychosocial characteristics. Finally, the modest analytic sample limited the number of covariates that could be included without reducing statistical precision.
Future research should employ longitudinal and experimental designs to better understand the causal mechanisms underlying the observed relationships. Additional studies involving more diverse populations are needed to determine whether the buffering role of AI trust extends across demographic, cultural, and educational contexts. Qualitative focus group discussions were conducted as a separate component of the broader research project. These data were not included in the present analysis and will be reported in a separate manuscript to provide a more in-depth understanding of how AI trust, stress, and resilience are experienced among students attending a Historically Black College or University (HBCU). These qualitative findings may help identify contextual factors that are not fully captured through survey-based measures. Future investigations should also examine specific forms of AI engagement, including digital mental health applications and AI-assisted decision-support tools, to identify the mechanisms through which trust influences resilience and well-being [20,21,28,33,34,36].

6. Conclusions

This study suggests that trust in artificial intelligence may be an important contextual correlate of how students experience the relationship between stress and mental toughness. Although AI trust did not directly predict mental toughness after accounting for the interaction effect, it significantly moderated the association between stress and mental toughness, with a weaker negative relationship observed at higher levels of AI trust. Because actual AI use was not measured, these findings should not be interpreted as evidence that AI technologies themselves improve resilience.
As AI continues to expand across healthcare and educational settings, public health efforts should prioritize transparency, ethical governance, fairness, and user-centered design to foster appropriate trust in AI-supported interventions. Future research should evaluate whether engagement with trustworthy AI-enabled mental wellness tools can support resilience, expand access to support services, and promote mental well-being among students and other populations experiencing elevated stress [22,24,28].

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/tph1020015/s1, Supplementary File S1: Student Well-Being and AI Experience Survey Instrument.

Author Contributions

Conceptualization, A.A. and K.B.; methodology, A.A. and V.K.; formal analysis, A.A. and V.K.; investigation, A.A., V.K., and K.B.; data curation, A.A.; writing—original draft preparation, A.A.; writing—review and editing, A.A., V.K., and K.B.; visualization, A.A. and V.K.; supervision, A.A.; project administration, A.A.; funding acquisition, A.A. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Transform Morgan 2030 Grant Program, Morgan State University (grant number: not applicable). The article processing charge (APC) was funded by the Transform Morgan 2030 Grant Program, Morgan State University.

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Morgan State University (IRB Protocol #26/02-0078; approved 25 February 2026).

Informed Consent Statement

Informed consent was obtained electronically from all subjects involved in the study.

Data Availability Statement

The data supporting the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy and ethical restrictions related to human participant research.

Acknowledgments

The authors extend their sincere appreciation to all students who participated in this study. We thank them for their openness, honesty, and thoughtful reflections regarding their well-being, experiences with stress, and comfort with the use of artificial intelligence. Their willingness to share their perspectives provided valuable insights that contributed significantly to this research. The authors also acknowledge the support of Morgan State University and the Transform Morgan 2030 initiative for fostering interdisciplinary research focused on student success, resilience, and well-being.

Conflicts of Interest

Akanksha Anand is the inventor of a U.S. provisional patent application (Application No. 64/050,373) related to AI-enabled mental wellness technologies. The remaining authors declare no conflicts of interest. The funders had no role in the study design, data collection, data analysis, manuscript preparation, or decision to publish the results.

Abbreviations

The following abbreviations are used in this manuscript:
AI Artificial Intelligence
APCArticle Processing Charge
DASS-21Depression Anxiety Stress Scales–21
HBCUHistorically Black College or University
IRBInstitutional Review Board
MTMental Toughness
TAMTechnology Acceptance Model

References

  1. American College Health Association. National College Health Assessment III: Undergraduate Student Reference Group Executive Summary; American College Health Association: Silver Spring, MD, USA, 2023. [Google Scholar]
  2. Pascoe, M.C.; Hetrick, S.E.; Parker, A.G. The impact of stress on students in higher education. Int. J. Adolesc. Youth 2020, 25, 104–112. [Google Scholar] [CrossRef] [Scilit]
  3. Lazarus, R.S.; Folkman, S. Stress, Appraisal, and Coping; Springer: New York, NY, USA, 1984. [Google Scholar]
  4. Antony, M.M.; Bieling, P.J.; Cox, B.J.; Enns, M.W.; Swinson, R.P. Psychometric properties of the 42-item and 21-item versions of the Depression Anxiety Stress Scales in clinical groups and a community sample. Psychol. Assess. 1998, 10, 176–181. [Google Scholar] [CrossRef] [Scilit]
  5. Auerbach, R.P.; Mortier, P.; Bruffaerts, R.; Alonso, J.; Benjet, C.; Cuijpers, P.; Demyttenaere, K.; Ebert, D.D.; Green, J.G.; Hasking, P.; et al. WHO World Mental Health Surveys International College Student Project: Prevalence and Distribution of Mental Disorders. J. Abnorm. Psychol. 2018, 127, 623–638. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  6. Lin, S.-H.; Huang, Y.-C. Life Stress and Academic Burnout. Act. Learn. High. Educ. 2014, 15, 77–90. [Google Scholar] [CrossRef] [Scilit]
  7. Ramadan, O.M.E.; Alruwaili, M.M.; Alruwaili, A.N.; Elsharkawy, N.B.; Abdelaziz, E.M.; El Badawy Ezzat, R.E.S.; El-Nasr, E.M.S. Digital Dilemma of Cyberbullying Victimization among High School Students: Prevalence, Risk Factors, and Associations with Stress and Mental Well-Being. Children 2024, 11, 634. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  8. Abulfaraj, G.G.; Upsher, R.; Zavos, H.M.S.; Dommett, E.J. The Impact of Resilience Interventions on University Students’ Mental Health and Well-Being: A Systematic Review. Educ. Sci. 2024, 14, 510. [Google Scholar] [CrossRef] [Scilit]
  9. Nogueira, M.J.C.; Sequeira, C.A. Positive and Negative Correlates of Psychological Well-Being and Distress in College Students’ Mental Health: A Correlational Study. Healthcare 2024, 12, 1085. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  10. Clough, P.; Strycharczyk, D.; Earle, K. Developing Mental Toughness: Improving Performance, Wellbeing and Positive Behaviour in Others; Kogan Page: London, UK, 2016. [Google Scholar]
  11. Gucciardi, D.F. Mental toughness: Progress and prospects. Curr. Opin. Psychol. 2017, 16, 17–23. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  12. Stamatis, A.; Morgan, G.B.; Boolani, A.; Papadakis, Z. The Positive Association between Grit and Mental Toughness, Enhanced by a Minimum of 75 Minutes of Moderate-to-Vigorous Physical Activity, among US Students. Psych 2024, 6, 221–235. [Google Scholar] [CrossRef] [Scilit]
  13. Stamatis, A.; Morgan, G.B.; Cowden, R.G.; Koutakis, P. Conceptualizing, measuring, and training mental toughness in sport: Perspectives of master strength and conditioning coaches. J. Study Sports Athl. Educ. 2023, 17, 128–145. [Google Scholar] [CrossRef] [Scilit]
  14. Glikson, E.; Woolley, A.W. Human trust in artificial intelligence: Review of empirical research. Acad. Manag. Ann. 2020, 14, 627–660. [Google Scholar] [CrossRef] [Scilit]
  15. Topol, E. Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again; Basic Books: New York, NY, USA, 2019. [Google Scholar]
  16. Davenport, T.H.; Ronanki, R. Artificial intelligence for the real world. Harv. Bus. Rev. 2018, 96, 108–116. [Google Scholar]
  17. Dwivedi, Y.K.; Kshetri, N.; Hughes, L.; Slade, E.L.; Jeyaraj, A.; Kar, A.K.; Baabdullah, A.M.; Koohang, A.; Raghavan, V.; Ahuja, M.; et al. So, what if ChatGPT wrote it? Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy. Int. J. Inf. Manag. 2023, 71, 102642. [Google Scholar] [CrossRef] [Scilit]
  18. Shneiderman, B. Human-centered artificial intelligence: Reliable, safe and trustworthy. Int. J. Hum. Comput. Interact. 2020, 36, 495–504. [Google Scholar] [CrossRef] [Scilit]
  19. Casu, M.; Triscari, S.; Battiato, S.; Guarnera, L.; Caponnetto, P. AI Chatbots for Mental Health: A Scoping Review of Effectiveness, Feasibility, and Applications. Appl. Sci. 2024, 14, 5889. [Google Scholar] [CrossRef] [Scilit]
  20. Fanarioti, A.K.; Karpouzis, K. Artificial Intelligence and the Future of Mental Health in a Digitally Transformed World. Computers 2025, 14, 259. [Google Scholar] [CrossRef] [Scilit]
  21. Ni, Y.; Jia, F. A Scoping Review of AI-Driven Digital Interventions in Mental Health Care: Mapping Applications Across Screening, Support, Monitoring, Prevention, and Clinical Education. Healthcare 2025, 13, 1205. [Google Scholar] [CrossRef] [Scilit]
  22. Jobin, A.; Ienca, M.; Vayena, E. The global landscape of AI ethics guidelines. Nat. Mach. Intell. 2019, 1, 389–399. [Google Scholar] [CrossRef] [Scilit]
  23. Floridi, L.; Cowls, J. A Unified Framework of Five Principles for AI in Society. Harv. Data Sci. Rev. 2019, 1. [Google Scholar] [CrossRef] [Scilit]
  24. World Health Organization. Ethics and Governance of Artificial Intelligence for Health; World Health Organization: Geneva, Switzerland, 2021. [Google Scholar]
  25. Frey, C.B.; Osborne, M.A. The future of employment: How susceptible are jobs to computerisation? Technol. Forecast. Soc. Change 2017, 114, 254–280. [Google Scholar] [CrossRef] [Scilit]
  26. Li, Y.; Wu, B.; Huang, Y.; Luan, S. Developing Trustworthy Artificial Intelligence: Insights from Research on Interpersonal, Human–Automation, and Human–AI Trust. Front. Psychol. 2024, 15, 1382693. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  27. Liu, F.; Ju, Q.; Zheng, Q.; Peng, Y. Artificial intelligence in mental health: Innovations brought by artificial intelligence techniques in stress detection and interventions of building resilience. Curr. Opin. Behav. Sci. 2024, 60, 101452. [Google Scholar] [CrossRef] [Scilit]
  28. Alharbi, B.S.; Aljabri, M.M.; Ali, E.A. Determinants of Trust in Artificial Intelligence (AI) for Health-Related Decision-Making Among Adults in Saudi Arabia: A Cross-Sectional Study. Healthcare 2026, 14, 506. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  29. Raisch, S.; Krakowski, S. Artificial intelligence and management: The automation–augmentation paradox. Acad. Manag. Rev. 2021, 46, 192–210. [Google Scholar] [CrossRef] [Scilit]
  30. Na, S.; Heo, S.; Han, S.; Shin, Y.; Roh, Y. Acceptance model of artificial intelligence (AI)-based technologies in construction firms: Applying the Technology Acceptance Model (TAM) in combination with the Technology–Organisation–Environment (TOE) framework. Buildings 2022, 12, 90. [Google Scholar] [CrossRef] [Scilit]
  31. Alsyouf, A.; Lutfi, A.; Alsubahi, N.; Alhazmi, F.N.; Al-Mugheed, K.; Anshasi, R.J.; Alharbi, N.I.; Albugami, M. The use of a Technology Acceptance Model (TAM) to predict patients’ usage of a personal health record system: The role of security, privacy, and usability. Int. J. Environ. Res. Public Health 2023, 20, 1347. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  32. Longoni, C.; Bonezzi, A.; Morewedge, C.K. Resistance to medical artificial intelligence. J. Consum. Res. 2019, 46, 629–650. [Google Scholar] [CrossRef] [Scilit]
  33. Balcombe, L. Digital Mental Health Post COVID-19: The Era of AI Chatbots. Encyclopedia 2026, 6, 32. [Google Scholar] [CrossRef] [Scilit]
  34. Balcombe, L. AI Chatbots in Digital Mental Health. Informatics 2023, 10, 82. [Google Scholar] [CrossRef] [Scilit]
  35. Davis, F.D. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q. 1989, 13, 319–340. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  36. Varghese, M.A.; Sharma, P.; Patwardhan, M. Public Perception on Artificial Intelligence–Driven Mental Health Interventions: Survey Research. JMIR Form. Res. 2024, 8, e64380. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  37. Nunnally, J.C.; Bernstein, I.H. Psychometric Theory, 3rd ed.; McGraw-Hill: New York, NY, USA, 1994. [Google Scholar]
  38. Venkatesh, V.; Morris, M.G.; Davis, G.B.; Davis, F.D. User acceptance of information technology: Toward a unified view. MIS Q. 2003, 27, 425–478. [Google Scholar] [CrossRef] [Scilit]
  39. Zakai, J.G.; Alharthi, S.A. Harnessing Digital Phenotyping for Early Self-Detection of Psychological Distress. Healthcare 2025, 13, 2008. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  40. Lindong, I.; Edwards, L.; Dennis, S.; Fajobi, O. Similarities and differences matter: Considering the influence of gender on HIV prevention programs for young adults in an urban HBCU. Int. J. Environ. Res. Public Health 2017, 14, 133. [Google Scholar] [CrossRef] [Scilit] [PubMed]
Tph 01 00015 g001
Figure 1. Moderating effect of AI trust on the relationship between stress and mental toughness. Higher levels of AI trust weakened the negative association between stress and mental toughness, suggesting a buffering pattern.
Figure 1. Moderating effect of AI trust on the relationship between stress and mental toughness. Higher levels of AI trust weakened the negative association between stress and mental toughness, suggesting a buffering pattern.
Tph 01 00015 g001
Table 1. Demographic characteristics of participants (n = 109).
Table 1. Demographic characteristics of participants (n = 109).
Characteristicn%
Age Group
18–25 years9284.4
26–30 years1110.1
31 years or older65.5
Gender Identity
Male3229.4
Female7568.8
Prefer not to say21.8
Academic Level
First-year undergraduate54.6
Sophomore65.5
Junior2321.1
Senior5954.1
Graduate student1614.7
Field of Study
STEM disciplines8174.3
Non-STEM disciplines2825.7
Employment Status
Employed full-time3229.4
Employed part-time3229.4
Not employed4541.3
Note: STEM = science, technology, engineering, and mathematics. Percentages may not total 100% because of rounding.
Table 2. Means, standard deviations, and correlations among study variables (n = 103).
Table 2. Means, standard deviations, and correlations among study variables (n = 103).
VariableMeanSD123
1. Stress (centered)−0.023.68—
2. AI Trust (centered)−0.002.98−0.006—
3. Mental Toughness19.853.42−0.340 ***0.187 *—
Note: Values represent Pearson correlation coefficients. * p < 0.05; *** p < 0.001.
Table 3. Hierarchical regression analysis predicting mental toughness (n = 103).
Table 3. Hierarchical regression analysis predicting mental toughness (n = 103).
PredictorModel 1 βp-ValueModel 2 βp-Value95% CI
Stress (centered)−0.339<0.001−0.326<0.001−0.470, −0.135
AI Trust (centered)0.1850.0480.1300.172−0.066, 0.363
Stress × AI Trust——0.2010.0360.003, 0.095
R20.150 0.187
ΔR2— 0.0370.036
Note: Standardized beta coefficients (β) are reported. Confidence intervals are reported for the unstandardized regression coefficients. All predictors were mean-centered prior to analysis.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Anand, A.; Kumar, V.; Bista, K. Artificial Intelligence Trust as a Buffer Against Stress: Implications for Mental Toughness and Student Well-Being. Trends Public Health 2026, 1, 15. https://doi.org/10.3390/tph1020015

AMA Style

Anand A, Kumar V, Bista K. Artificial Intelligence Trust as a Buffer Against Stress: Implications for Mental Toughness and Student Well-Being. Trends in Public Health. 2026; 1(2):15. https://doi.org/10.3390/tph1020015

Chicago/Turabian Style

Anand, Akanksha, Vishnu Kumar, and Krishna Bista. 2026. "Artificial Intelligence Trust as a Buffer Against Stress: Implications for Mental Toughness and Student Well-Being" Trends in Public Health 1, no. 2: 15. https://doi.org/10.3390/tph1020015

APA Style

Anand, A., Kumar, V., & Bista, K. (2026). Artificial Intelligence Trust as a Buffer Against Stress: Implications for Mental Toughness and Student Well-Being. Trends in Public Health, 1(2), 15. https://doi.org/10.3390/tph1020015

Article Metrics

Back to TopTop