The potential of artificial intelligence in psychological profiling according to the subjective assessments of psychologists
DOI:
https://doi.org/10.17721/BPSY.2026.1(23).13Keywords:
AI, large language models, digital traces, psychodiagnostics, profilingAbstract
Background. In the context of the full-scale Russian-Ukrainian war and the growing need for rapid psychological assessment of security and defense personnel, the search for new psychodiagnostic tools has become especially relevant. One promising area is the use of artificial intelligence (AI) for psychological profiling based on users' digital traces, particularly the history of their interactions with large language models. The aim of the study was to assess, based on a survey of psychology specialists, the quality and accuracy of personality profiles generated by AI systems.
Methods. The study was conducted as an anonymous online survey with elements of a natural experiment. Participants independently ran standardized prompts in the AI systems they normally use, which had access to the previous interaction context, and then evaluated the accuracy of the generated profiles. The final analysis included 29 psychology specialists. Three approaches were assessed: the Big Five (OCEAN), Nancy McWilliams' psychodynamic model, and analytical profiling. Data were processed using descriptive statistics, the Friedman and Wilcoxon tests, and Pearson correlation analysis.
Results. All three approaches received similar accuracy ratings within a moderately good range of correspondence to reality: Big Five – 7,1 points, psychodynamic approach – 6,9 points, and analytical profiling – 7,2 points on a 10-point scale. The overall subjective rating of AI profiling quality was also 7,1 points. No statistically significant differences were found between the approaches. It was established that specialists with higher self-rated competence in profiling evaluated AI results more critically, whereas greater thematic diversity in prior interactions with AI was associated with higher accuracy ratings. Qualitative responses from participants additionally showed that the effectiveness of AI profiling depends substantially on the amount and substantive diversity of the available context.
Conclusions. Modern artificial intelligence systems are capable of generating psychological profiles that specialists evaluate as moderately well corresponding to reality. The results indicate the promise of using AI as a supportive tool in psychological assessment, particularly in the psychological selection and support of security and defense personnel. At the same time, the use of such systems requires a critical attitude toward the results, further objective validation, and careful consideration of personal data protection issues.
Downloads
References
., & Семененко, А. (2025). Військова педагогіка та психологія в умовах російсько-української війни: проблеми, рішення і перспективи розвитку. Міжнародний науковий журнал "Military Science", 3(3), 66-82. https://doi.org/10.62524/msj.2025.3.3.5
Eichstaedt, J. C., Smith, R. J., Merchant, R. M., Ungar, L. H., Crutchley, P., Preoţiuc-Pietro, D., Asch, D. A., & Schwartz, H. A. (2018). Facebook language predicts depression in medical records. Proceedings of the National Academy of Sciences, 115(44), 11203–11208. https://doi.org/10.1073/pnas.1802331115
Guntuku, S. C., Yaden, D. B., Kern, M. L., Ungar, L. H., & Eichstaedt, J. C. (2017). Detecting depression and mental illness on social media: An integrative review. Current Opinion in Behavioral Sciences, 18, 43–49. https://doi.org/10.1016/j.cobeha.2017.07.005
Kosinski, M., Stillwell, D., & Graepel, T. (2013). Private traits and attributes are predictable from digital records of human behavior. Proceedings of the National Academy of Sciences, 110(15), 5802–5805. https://doi.org/10.1073/pnas.1218772110
Lingiardi, V., & McWilliams, N. (Eds.). (2026). Psychodynamic Diagnostic Manual (3rd ed.; PDM-3). Guilford Press. ISBN 9781462558711.
Matz, S. C., Kosinski, M., Nave, G., & Stillwell, D. J. (2017). Psychological targeting as an effective approach to digital mass persuasion. Proceedings of the National Academy of Sciences, 114(48), 12714–12719. https://doi.org/10.1073/pnas.1710966114
Ocana Flores, H. I., & Luna, A. (2024). AI for psychological profiles: Advances, challenges, and future directions. Ciencia Latina Revista Científica Multidisciplinar, 8(3), 10592–10609. https://doi.org/10.37811/cl_rcm.v8i3.12221
Papneja, H., & Yadav, N. (2025). Self-disclosure to conversational AI: A literature review, emergent framework, and directions for future research. Personal and Ubiquitous Computing, 29, 119–151. https://doi.org/10.1007/s00779-024-01823-7
Pataranutaporn, P., Lee, E., Amores, J., & Maes, P. (2026). Personal validation effect in LLMs: Positive AI responses bias perceptions of validity, reliability, personalization, and usefulness of fictitious predictions. In Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (pp. 1–18). ACM. https://doi.org/10.1145/3772318.3791851
Peters, H., Cerf, M., & Matz, S. C. (2024). Large language models can infer personality from free-form user interactions (arXiv:2405.13052). arXiv. https://doi.org/10.48550/arXiv.2405.13052
Staab, R., Vero, M., Balunović, M., & Vechev, M. (2024). Beyond memorization: Violating privacy via inference with large language models. In Proceedings of the Twelfth International Conference on Learning Representations (ICLR 2024). https://doi.org/10.48550/arXiv.2310.07298
Föyen, L. F., Zapel, E., Lekander, M., Hedman-Lagerlöf, E., & Lindsäter, E. (2025). Artificial intelligence vs. human expert: Licensed mental health clinicians' blinded evaluation of AI-generated and expert psychological advice on quality, empathy, and perceived authorship. Internet Interventions. https://doi.org/10.1016/j.invent.2025.100841
Youyou, W., Kosinski, M., & Stillwell, D. (2015). Computer-based personality judgments are more accurate than those made by humans. Proceedings of the National Academy of Sciences, 112(4), 1036–1040. https://doi.org/10.1073/pnas.1418680112
Zell, E., & Lesick, T. L. (2022). Big five personality traits and performance: A quantitative synthesis of 50+ meta-analyses. Journal of Personality, 90(4), 559–573. https://doi.org/10.1111/jopy.12683
Zhu, J., Jin, R., & Coifman, K. G. (2025). Investigating large language models in inferring personality traits from user conversations (arXiv:2501.07532). arXiv. https://doi.org/10.48550/arXiv.2501.07532
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Bulletin of Taras Shevchenko National University of Kyiv. Psychology
This work is licensed under a Creative Commons Attribution 4.0 International License.
