AI Work Group Recommendations

The American Counseling Association has convened a panel of counseling experts representing academia, private practice and students to comprise its AI Work Group. The work group used research-based and contextual evidence; the ACA Code of Ethics; and clinical knowledge and skill to develop the following recommendations. The goal is to both prioritize client well-being, preferences, and values in the advent and application of AI, while informing counselors, counselor-educators and clients about the use of AI today. The recommendations also highlight the additional research needed to inform counseling practice as AI becomes a more widely available and accepted part of mental health care.

While research into the role of Artificial Intelligence (AI) in assessment and diagnosis has expanded, more studies are needed to explore AI's assistance to counselors in these domains. The ensuing recommendations adopt an interdisciplinary approach to AI's integration into assessment and diagnosis processes, highlighting potential applications and encouraging caution. Counselors are urged to pursue additional research in this field and closely track AI advancements. Counselors are advised to exercise prudence when leveraging AI support for assessment and diagnosis while also maintaining receptivity to its possibilities.

Recommendation: AI may augment or help reimagine diagnostic frameworks
Recognize that AI may lead counselors to reconsider how they categorize mental health disorders (Minerva & Giubilini, 2023). At present, counselors diagnose by observing, classifying, and assessing predominately in accordance to their level of ability. Counselors may consult with others to improve the validity of their diagnosis. AI may expand a counselor's knowledge base by including speech-pattern analysis and other datasets, providing more markers correlating with specific diagnoses. In this sense, AI would help fine-tune diagnosis, which may render DSM-style categorical diagnoses incomplete or in need of substantial revision.

Recommendation: Carefully examine AI outputs for potential bias
Recognize that AI may reduce or increase the incidence of bias in diagnosis. Mainly depending on the dataset in which it is trained, the diagnosis offered by an AI may be more or less biased than a human counselor. AI trained with representative datasets may offer a degree of impartiality to diagnosis, serving as a valid source of information for the counselor to consider before giving a diagnosis. Conversely, AI trained with non-representative datasets may incorrectly diagnose, miss a diagnosis, or lead the counselor astray with miscalculated information. The process mirrors Type I and II errors in statistics. Counselors are encouraged to scrutinize the offerings of AI for possible bias (Fulmer et al., 2021)

Recommendation: Weighing AI's potential against ethical obligations is imperative
Integrating AI into assessment may help inform the diagnostic process, aiding counselors in improving the accuracy, consistency, and objectivity of diagnoses, given the potential for humans to be influenced by their own emotions and cognitive biases when conducting assessment and issuing diagnoses (Featherston et al., 2020). Despite industry assurances, it is imperative to check whether the AI-powered applications and software for clinical purposes comply with HIPAA, local laws, and employer or clinical site regulations and policies. Data can easily be reidentified in the digital era (Marks & Haupt, 2023). More research and legislation efforts are needed to establish HIPAA-compliant LLMs and other AI-power applications with specific knowledge and considerations pertinent to the counseling profession. Practicing counselors and counseling researchers are encouraged to actively participate in developing these LLMs and other machine-learning models in light of their domain knowledge and practical experience.

Recommendation: Ensure HIPAA compliance
Session notes help with assessment and diagnosis in counseling (Prieto & Scheel, 2002). Despite the temptation to use AI tools for automated note-taking, client data should not be input into non-HIPAA-compliant applications. Failure to do so will result in a violation of HIPAA laws and confidentiality. Before adopting AI tools, counselors must practice due diligence and ensure these tools comply with HIPAA and local laws and regulations. If using HIPAA-compliant AI tools for automated note-taking, counselors must review notes for accuracy and adherence to professional standards and edit automated notes for compliance with professional standards as needed.

Recommendation: Develop ethical, accurate AI tools through diverse data sets and collaboration
To develop AI tools that are ethical, just, accurate, and efficient in counseling, research should incorporate inclusive and diverse data sets that reflect a broad spectrum of client experiences to reduce bias and improve the generalizability of the algorithms. This effort should be underpinned by interdisciplinary collaboration, bringing together counseling and computer science researchers, counselors, and clients to integrate clinical expertise with AI development, ensuring the tools are clinically relevant and grounded in therapeutic best practices. Rigorous counseling research can help examine the role of AI tools in counseling, from targeted prevention and early intervention to assessment and diagnosis. Their function can be compared with traditional approaches to test their effectiveness, accuracy, efficiency, and client satisfaction. Engaging clients in the research process to gather feedback on their experiences with AI tools can also help enhance the development process and understand client perception, paving the way for client-centered AI tools that are more aligned with client needs and expectations.

Recommendation: Exercise caution and critical thinking
Counselors using AI to facilitate assessment and diagnosis are encouraged to exercise caution, critical thinking, and remember not to rely solely on the information an AI provides. In accordance with ethics code E.9.b., counselors should “qualify any conclusions, diagnoses, or recommendations made that are based on assessments or instruments with questionable validity or reliability.”

Recommendation: Counselors should integrate AI insights with relational dynamics for cultural responsiveness
As indicated in the ACA Code of Ethics (ACA, 2014), “Counselors use assessment as one component of the counseling process, taking into account the clients’ personal and cultural context.” (Section E, Introduction). AI tools may potentially be a useful aid for counselors seeking to understand broad cultural considerations of clients. However, counselors must ensure that they are attending to relational dynamics with clients when exploring cultural formulations of presenting problems to promote individualized therapeutic relationships and culturally responsive treatment planning


Abd-Alrazaq, A., Alhuwail, D., Schneider, J., Toro, C. T., Ahmed, A., Alzubaidi, M., ... & Househ, M. (2022). The performance of artificial intelligence-driven technologies in diagnosing mental disorders: an umbrella review. NPJ Digital Medicine, 5(1), 87.

Featherston, R., Downie, L. E., Vogel, A. P., & Galvin, K. L. (2020). Decision making biases in the allied health professions: a systematic scoping review. PLoS One, 15(10), e0240716.

Jarvis, G. E., Kirmayer, L. J., Gómez-Carrillo, A., Aggarwal, N. K., & Lewis-Fernández, R. (2020). Update on the cultural formulation interview. Focus, 18(1), 40-46.

Kulkarni, P. A., & Singh, H. (2023). Artificial Intelligence in Clinical Diagnosis: Opportunities, Challenges, and Hype. JAMA, 330(4), 317–318.

Lewis-Fernández, R., Aggarwal, N. K., & Kirmayer, L. J. (2020). The Cultural Formulation Interview: Progress to date and future directions. Transcultural Psychiatry, 57(4), 487-496.

Lewis-Fernández, R., Aggarwal, N. K., Lam, P. C., Galfalvy, H., Weiss, M. G., Kirmayer, L. J., ... & Vega-Dienstmaier, J. M. (2017). Feasibility, acceptability and clinical utility of the Cultural Formulation Interview: mixed-methods results from the DSM-5 international field trial. The British Journal of Psychiatry, 210(4), 290-297.

Fulmer, R., Davis, T., Costello, C., & Joerin, A. (2021). The ethics of psychological artificial intelligence: Clinical considerations. Counseling and Values, 66(2), 131-144.

Marks, M., & Haupt, C. E. (2023). AI Chatbots, health privacy, and challenges to HIPAA compliance. Jama.

Minerva, F., & Giubilini, A. (2023). Is AI the Future of Mental Healthcare?. Topoi : an international review of philosophy, 42(3), 1–9. Advance online publication.

Prieto, L. R., & Scheel, K. R. (2002). Using case documentation to strengthen counselor trainees' case conceptualization skills. Journal of Counseling & Development, 80(1), 11-21.

Sun, J., Dong, Q.-X., Wang, S.-W., Zheng, Y.-B., Liu, X.-X., Lu, T.-S., Yuan, K., Shi, J., Hu, B., Lu, L., & Han, Y. (2023). Artificial intelligence in psychiatry research, diagnosis, and therapy. Asian Journal of Psychiatry, 87.

AI Work Group Members

S. Kent Butler, PhD
University of Central Florida
Russell Fulmer, PhD
Husson University
Morgan Stohlman
Kent State University
Fallon Calandriello, PhD
Northwestern University
Marcelle Giovannetti, EdD
Messiah University- Mechanicsburg, PA
Olivia Uwamahoro Williams, PhD
College of William and Mary
Wendell Callahan, PhD
University of San Diego
Marty Jencius, PhD
Kent State University
Yusen Zhai, PhD
UAB School of Education
Lauren Epshteyn
Northwestern University
Sidney Shaw, EdD
Walden University
Chip Flater
Dania Fakhro, PhD
University of North Carolina, Charlotte