Readiness evaluation for artificial intelligence-mental health deployment and implementation (READI): A review and proposed framework.
Open accessPublished version of a preprint
Abstract
While generative artificial intelligence (AI) may lead to technological advances in the mental health field, it poses safety risks for mental health service consumers. Furthermore, clinicians and healthcare systems must attend to safety and ethical considerations prior to deploying these AI-mental health technologies. To ensure the responsible deployment of AI-mental health applications, a principled method for evaluating and reporting on AI-mental health applications is needed. We conducted a narrative review of existing frameworks and criteria (from the mental health, healthcare, and AI fields) relevant to the evaluation of AI-mental health applications. We provide a summary and analysis of these frameworks, with a particular emphasis on the unique needs of the AI-mental health intersection. Existing frameworks contain areas of convergence (e.g., frequent emphasis on safety, privacy/confidentiality, effectiveness, and equity) that are relevant to the evaluation of AI-mental health applications. However, current frameworks are insufficiently tailored to unique considerations for AI and mental health. To address this need, we introduce the Readiness Evaluation for AI-Mental Health Deployment and Implementation (READI) framework for mental health applications. The READI framework comprises considerations of Safety, Privacy/Confidentiality, Equity, Effectiveness, Engagement, and Implementation. The READI framework outlines key criteria for assessing the readiness of AI-mental health applications for clinical deployment, offering a structured approach for evaluating these technologies and reporting findings.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
LLM-CBs find increasing use as mental health support tools, both for the general public and for patients in clinical interventions.
“LLM-CBs find increasing use as mental health support tools, both for the general public8–12 and for patients in clinical interventions13,14.”
From A scoping review on the mental health harms of LLM-based chatbots (2026) · cites this paper as [11]
The Results section describes patient-facing psychosocial interventions as a prospective use of AI in mental healthcare: applications are “poised” to deliver them. That supports the clinical-intervention context, but not the claim that LLM chatbots are already seeing increasing use by patients in those interventions. The cited passage has no page number in our copy. (LLM, conf=0.84)
Experimental — accuracy not yet measuredSurveys show that the rates of individuals using LLMs for mental health range from 24% to 48.7% of users.
“Surveys show that the rates of individuals using LLMs for mental health range from 24% to 48.7% of users8–12.”
From A scoping review on the mental health harms of LLM-based chatbots (2026) · cites this paper as [11]
The Abstract discusses potential mental-health applications of LLMs, and the Results section discusses evaluating engagement, but neither reports survey rates of use. Searches for “24%,” “48.7%,” and survey prevalence found no passage supporting the claimed range in our copy; page numbers were not provided. (LLM, conf=0.76)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
The publisher hasn’t deposited this paper’s reference list.
Research works printed but not linked: 1
- Beck AT (1987). Cognitive models of depression. Journal of Cognitive Psychotherapy: An International Quarterly, 1, 5– 37.
Cited by
Sorted by citations
Scholich, T., Barr, M., Wiltsey Stirman, S., & Raj, S. (2025) · JMIR Mental Health · doi:10.2196/69709 · Cited by 72 · see citing works
Voultsiou, E., & Moussiades, L. (2026) · Electronics · doi:10.3390/electronics15030524 · Cited by 16 · see citing works
Artificial Intelligence in Obsessive-Compulsive Disorder: A Systematic Review
Kim, J. Y., Pacheco, J. P. G., Golden, A., Aboujaoude, E. N., van Roessel, P., Gandhi, A., Mukunda, P., Avanesyan, T., Xue, H., Adeli, E., et al. (2025) · Current Treatment Options in Psychiatry · doi:10.1007/s40501-025-00359-8 · Cited by 12 · see citing works
Self-citationHeld, P., Stade, E. C., Dondanville, K. A., & Stirman, S. W. (2025) · Journal of Traumatic Stress · doi:10.1002/jts.23188 · Cited by 8 · see citing works
Self-citationShafran, R., Bond, L., Carlbring, P., Cohen, Z. D., Creed, T. A., Davey, E., Egan, S. J., Freeman, D. H., Hollon, S. D., Jacobson, N. C., et al. (2025) · Behaviour Research and Therapy · doi:10.1016/j.brat.2025.104945 · Cited by 10 · see citing works
Sawesi, S., Sabbineni, H., Shagamreddy, R. R., & Rashrash, B. (2026) · Journal of Multidisciplinary Healthcare · doi:10.2147/jmdh.s581251 · Cited by 5 · see citing works
A scoping review on the mental health harms of LLM-based chatbots
Diel, A., Torous, J., Cuijpers, P., Kleesiek, J., Nensa, F., Weber, N., Faust, F., Lalgi, T. J., Mellis, F. S., Teufel, M., et al. (2026) · npj Digital Medicine · doi:10.1038/s41746-026-03054-x · Cited by 4 · see citing works
ElBarazi, A. S., Mohamed, H., & Nasser, R. (2026) · Healthcare · doi:10.3390/healthcare14091247 · Cited by 5 · see citing works
Efficiency Pitfalls of Explainable AI in Clinical Diagnostic and Treatment Human-AI Workflows
Hunsicker, T., Schulz, A., Leist, R. A., Kiefer, S., Boden, K., Rothaus, K., König, C. J., & Langer, M. (2026) · Human Factors The Journal of the Human Factors and Ergonomics Society · doi:10.1177/00187208261443764 · Cited by 2 · see citing works
Exploring Real-World Use of AI Chatbots for Mental Health Support: Cross-Sectional Survey Study
Bernaerts, S., Buelens, F., Witte, N. A. J. D., Balaskas, A., Colman, T., & Daele, T. V. (2026) · JMIR Mental Health · doi:10.2196/104316 · Cited by 2 · see citing works
Nguyen, M. V., Khanh, H. D., Thuc, L. D., & Nguyễn, T. T. (2025) · Engineering Construction & Architectural Management · doi:10.1108/ecam-02-2025-0294 · Cited by 1 · see citing works
A Risk Ontology for Evaluating AI-Powered Psychotherapy Virtual Agents
Steenstra, I., & Bickmore, T. (2025) · ACM International Conference on Intelligent Virtual Agents (IVA) · doi:10.1145/3717511.3749286 · Cited by 1 · see citing works
Developing mental health AI tools that improve care across different groups and contexts
Martinez‐Martin, N. (2025) · Nature Computational Science · doi:10.1038/s43588-025-00882-x · Cited by 1 · see citing works
Ridings, L. E., Held, P., Kuhn, E. R., & Stirman, S. W. (2026) · Journal of Traumatic Stress · doi:10.1002/jts.70092 · Cited by 1 · see citing works
Lee, J. L., Billovits, C., Chen, S.-Y., Wickham, R. E., Kocher, B., Chen, C., & Lungu, A. (2026) · Value in Health · doi:10.1016/j.jval.2025.11.023
Lange, M., Koutsouleris, N., Carter, B., & Twumasi, R. (2026) · Healthcare Technology Letters · doi:10.1049/htl2.70058
Clark, C., Jerskey, B. A., & Fogler, J. (2026) · Journal of Pediatric Psychology · doi:10.1093/jpepsy/jsag027
A Behavioral Health Practitioner’s Primer for the Use of Artificial Intelligence (AI)
Perle, J. G. (2026) · Journal of Technology in Behavioral Science · doi:10.1007/s41347-026-00636-0
King, A. J., Banks, A., Hernández, L. H., Thompson, S. S., Stevens, L., Potter, L. N., Kaphingst, K. A., Estabrooks, P. A., Fiol, G. D., Wetter, D. W., et al. (2026) · Journal of Medical Internet Research · doi:10.2196/92356
Hennemann, S., Fähnrich, J. M., Tietze, C., Witthöft, M., & Jungmann, S. M. (2026) · Digital Health · doi:10.1177/20552076261450336
Albikawi, Z. F., Abuadas, M. H., Saifan, A. R., & Ahmed, F. R. (2026) · IEEE Annual International Computer Software and Applications Conference (COMPSAC) · doi:10.1109/compsac69091.2026.00336
Effectiveness of Student-Centered Fine-Tuning of Large Language Models for Mental Health Support
Musmade, S., & Liu, L. (2026) · IEEE International Conference on Big Data Computing Service and Applications (BigDataService) · doi:10.1109/bigdataservice70481.2026.00024
Salinas-Layana, A., Jiménez-Molina, Á., Lira, D., Montoya, F. V., Venegas, A., Muñoz, N., Chandía, M., Rojas, R., & Martínez, V. (2026) · JMIR Research Protocols · doi:10.2196/91677
- Authors
- Stade, E. C., Eichstaedt, J. C., Kim, J. P., & Wiltsey Stirman, S.
- Affiliations
- Stanford University (US)
- Stanford Medicine (US)
- Venue
- Technology, Mind, and Behavior
- Published
- 2025
- Volume
- 6
- Issue
- 2
- Pages
- 111-122
- ISSN
- 2689-0208 (ISSN-L)
- PMC
- PMC12494062
- OpenAlex
- W4409767709
- Cited by
- 23
- References
- 0
- Citations by year
- 2026: 16 · 2025: 7
- FWCI
- 5.54
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Technology Mind and Behavior (publishedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- PubMed Central (publishedVersion)
- Funders
- National Institute of Mental Health (R01-MH125702, RF1-MH128785)
- National Center for Advancing Translational Sciences (R01-TR003505)
- Stanford University and Stanford Institute for Human-Centered Artificial Intelligence
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Keywords
- generative AI, evaluation framework, safety, privacy and confidentiality, equity, effectiveness, user engagement, implementation, ethical considerations, responsible AI deployment
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
