Balancing risks and benefits: clinicians’ perspectives on the use of generative AI chatbots in mental healthcare
Open accessThe DOI matches its Crossref record; 25 of the 25 references it cites were found in a registry or an index.
Abstract
Introduction: The use of generative-AI chatbots has proliferated in mental health, to support both clients and clinicians across a range of uses. This paper aimed to explore the perspectives of mental health clinicians regarding the risks and benefits of integrating generative-AI chatbots into the mental health landscape. Methods: Twenty-three clinicians participated in a 45-minute virtual interview, in which a series of open-ended and scale-based questions were asked, and a demonstration of a mental health chatbot's potential capabilities was presented. Results: Participants highlighted several benefits of chatbots, such as their ability to administer homework tasks, provide multilingual support, enhance accessibility and affordability of mental healthcare, offer access to up-to-date research, and increase engagement in some client groups. However, they also identified risks, including the lack of regulation, data and privacy concerns, chatbots' limited understanding of client backgrounds, potential for client over-reliance on chatbots, incorrect treatment recommendations, and the inability to detect subtle communication cues, such as tone and eye contact. There was no significant finding to suggest that participants viewed either the risks or benefits as outweighing the other. Moreover, a demonstration of potential chatbot capabilities was not found to influence whether participants favoured the risks or benefits of chatbots. Discussion: Qualitative responses revealed that the balance of risks and benefits is highly contextual, varying based on the use case and the population group being served. This study contributes important insights from critical stakeholders for chatbot developers to consider in future iterations of AI tools for mental health.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Survey of Hallucination in Natural Language Generation
Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., Ishii, E., Bang, Y. J., Madotto, A., & Fung, P. (2023) · ACM Computing Surveys · doi:10.1145/3571730 · Cited by 5,025 · cited 1×
“It also risks generating “ hallucinations, ” where the AI generates information that is incorrect, misleading, or entirely fabricated (25, 36).”
Survey of Hallucination in Natural Language Generation
Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., Ishii, E., Bang, Y., Chen, D., Dai, W., et al. (2022) · arXiv · doi:10.48550/arxiv.2202.03629 · Cited by 4,981
Torous, J., Bucci, S., Bell, I. H., Kessing, L. V., Faurholt‐Jepsen, M., Whelan, P., Carvalho, A. F., Keshavan, M., Linardon, J., & Firth, J. (2021) · World Psychiatry · doi:10.1002/wps.20883 · Cited by 1,199
Chatbots and Conversational Agents in Mental Health: A Review of the Psychiatric Landscape
Vaidyam, A. N., Wisniewski, H., Halamka, J. D., Kashavan, M. S., & Torous, J. B. (2019) · The Canadian Journal of Psychiatry · doi:10.1177/0706743719828977 · Cited by 1,048
An overview of the features of chatbots in mental health: A scoping review
Abd-alrazaq, A. A., Alajlani, M., Alalwan, A. A., Bewick, B. M., Gardner, P., & Househ, M. (2019) · International Journal of Medical Informatics · doi:10.1016/j.ijmedinf.2019.103978 · Cited by 605
Sharma, A., Lin, I. W., Miner, A. S., Atkins, D. C., & Althoff, T. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-022-00593-2 · Cited by 500
Accuracy of a Generative Artificial Intelligence Model in a Complex Diagnostic Challenge
Kanjee, Z., Crowe, B., & Rodman, A. (2023) · JAMA · doi:10.1001/jama.2023.8288 · Cited by 449
Haque, M. D. R., & Rubya, S. (2023) · JMIR mHealth and uHealth · doi:10.2196/44838 · Cited by 404 · cited 1×
“…handling of disclosures of criminal activity (29), the lack of comprehensive regulatory frameworks (30, 31), chatbots ’ inabilities to fully grasp the complexities of individual client situations (32, 33), and the risk of AI suggesting inappropriate diagnoses or treatment recommendations (34, 35).”
Liu, H., Peng, H., Song, X., Xu, C., & Zhang, M. (2022) · Internet Interventions · doi:10.1016/j.invent.2022.100495 · Cited by 296
Kretzschmar, K., Tyroll, H., Pavarini, G., Manzini, A., Singh, I., & NeurOx Young People’s Advisory Group (2019) · Biomedical Informatics Insights · doi:10.1177/1178222619829083 · Cited by 270
Research works printed but not linked: 20
- Xu, 2022, Survey on psychotherapy chatbots, Concurr Comput Pract Exp, 34, e6170, 10.1002/cpe.6170 · DOI · Paper page
- Cameron, 2019, Assessing the usability of a chatbot for mental health care, 121
- Jang, 2021, Mobile app-based chatbot to deliver cognitive behavioral therapy and psychoeducation for adults with attention deficit: a development and feasibility/usability study, Int J Med Inform, 150, 104440, 10.1016/j.ijmedinf.2021.104440 · DOI · Paper page
- Potts, 2023, A multilingual digital mental health and well-being chatbot (ChatPal): pre-post multicenter intervention study, J Med Internet Res, 25, e43051, 10.2196/43051 · DOI · Paper page
- Viduani, 2023, Chatbots in the field of mental health: challenges and opportunities, 133, 10.1007/978-3-031-10698-9_8 · DOI
- Suharwardy, 2023, Feasibility and impact of a mental health chatbot on postpartum mental health: a randomized controlled trial, AJOG Glob Rep, 3, 100165, 10.1016/j.xagr.2023.100165 · DOI · Paper page
- Maurya, 2024, Assessing the use of ChatGPT as a psychoeducational tool for mental health practice, Couns Psychother Res, 25, 1, 10.1002/capr.12759 · DOI · Paper page
- Yenduri, 2023, Generative pre-trained transformer: a comprehensive review on enabling technologies, potential applications, emerging challenges, and future directions, arXiv:230510435, 1, 10.48550/arXiv.2305.10435 · DOI
- Cabrera, 2023, Ethical dilemmas, mental health, artificial intelligence, and LLM-based chatbots, 313, 10.1007/978-3-031-34960-7_22 · DOI
- Templin, 2024, Addressing 6 challenges in generative AI for digital health: a scoping review, PLOS Digit Health, 3, e0000503, 10.1371/journal.pdig.0000503 · DOI
- Lee, 2021, Artificial intelligence for mental health care: clinical applications, barriers, facilitators, and artificial wisdom, Biol Psychiatry Cogn Neurosci Neuroimaging, 6, 856, 10.1016/j.bpsc.2021.02.001 · DOI · Paper page
- Stiefel, 2018, The chatbot will see you now': mental health confidentiality concerns in software therapy, SSRN, 1, 10.2139/ssrn.3166640 · doi:10.2139/ssrn.3166640
- Coghlan, 2023, To chat or bot to chat: ethical issues with using chatbots in mental health, Digit Health, 9, 20552076231183542, 10.1177/20552076231183542 · DOI · Paper page
- Balcombe, 2023, AI Chatbots in digital mental health, Informatics, 10, 82, 10.3390/informatics10040082 · DOI · Paper page
- Gallese, 2022, Legal issues of the use of chatbot apps for mental health support, 258
- Sweeney, 2021, Can chatbots help support a person’s mental health? Perceptions and views from mental healthcare professionals and experts, ACM Trans Comput Healthc, 2, 1, 10.1145/3453175 · DOI · Paper page
- Stroud, 2025, Physician perspectives on the potential benefits and risks of applying artificial intelligence in psychiatric medicine: qualitative study, JMIR Ment Health, 12, e64414, 10.2196/64414 · DOI
- Vo, 2023, Multi-stakeholder preferences for the use of artificial intelligence in healthcare: a systematic review and thematic analysis, Soc Sci Med, 338, 116357, 10.1016/j.socscimed.2023.116357 · DOI · Paper page
- Carifio, 2008, Resolving the 50-year debate around using and misusing Likert scales, Med Educ, 42, 1150, 10.1111/j.1365-2923.2008.03172.x · DOI · Paper page
- Wagenmakers, 2018, Bayesian Inference for psychology. Part I: theoretical advantages and practical ramifications, Psychon Bull Rev, 25, 35, 10.3758/s13423-017-1343-3 · DOI · Paper page
Cited by
Chung, V.-H.-A., Bernier, P., & Hudon, A. (2026) · JMIR Mental Health · doi:10.2196/93040 · Cited by 13
Kuhail, M. A., Al-Shamaileh, O., Farooq, S., Shahin, H., Abdelzaher, F., & Thomas, J. (2025) · Human Behavior and Emerging Technologies · doi:10.1155/hbe2/9942295 · Cited by 11
Shankar, R., Lim, A., & Xu, Q. (2026) · Healthcare · doi:10.3390/healthcare14101267 · Cited by 9
Unlocking AI Chatbot Potential in Healthcare: Trust-Enhanced DeLone & McLean IS Success Model
Sarhan, M. Y., Alarify, M., & Khojah, M. (2026) · Healthcare · doi:10.3390/healthcare14101324 · Cited by 2
“Others have gone so far as to demonstrate benefits and utilization of AI-powered chatbots in various healthcare environments [5,35,45–47], but how effectively it is utilized to provide appropriate service support to the users remains an unresolved concern in the extant literature.”
Metadata
- Authors
- Hipgrave, L., Goldie, J., Dennis, S., & Coleman, A.
- Venue
- Frontiers in Digital Health
- Publisher
- Frontiers Media SA
- Published
- 2025-05-29
- Volume
- 7
- Pages
- 1606291-1606291
- ISSN
- 2673-253X (ISSN-L)
- OpenAlex
- W4410872768
- Cited by
- 51
- References
- 25
- FWCI
- 37.01
- Type
- journal article
- Language
- en
- Open access
- gold
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.