AI in Child and Youth Mental Health Services: Policy Pathways for Ethical Practice

Abstract

The purpose of this study is to explore how Artificial Intelligence tools (AI) are being used in clinical settings that support the mental health of children and youth. As the use of AI tools become more widespread, exploring how clinicians perceive tools provide context into how AI can assist in managing work tasks despite concerns. Current literature highlights the benefits of AI tools in clinical settings in terms of decision making, assessment, and intervention. In contrast, current literature highlights concern around regulation/oversight and privacy. The study integrated a mixed methods design to gather the perception of clinicians and supervisors employed by MCFD's children and youth mental health (CYMH) service (N=42). A jurisdictional scan was also conducted to explore how other organizations similar to CYMH's mandate integrate AI-tools. The mixed methods design was chosen to gather an in-depth exploration of how clinicians perceive AI tools used in their work at CYMH. The Quantitative and Qualitative results are reported separately to form the results at the end of data collection to converge and identify core themes. There will be a discussion around the consideration of both datasets to explore the impact of results towards current knowledge around the use of AI-tools in CYMH. Based on findings, 69% (n=42) of respondents perceive AI tools being beneficial towards work tasks. In contrast, 88.1% of respondents perceive concerns around the ethical implications of AI-tool use in clinical practice. In addition, 83.5% (n=42) of participants agreed with receiving additional training and policies that guide the use of AI-tools in clinical settings with children and youth. Three major key themes were identified: (1) clinicians' uncertainties about using AI in clinical practice, (2) clinicians recognize that AI use in clinical practice offers advantages but also presents challenges and (3) clinicians are underprepared by current policy and training. Findings of the study align with the main themes of the jurisdictional scan, where policies place an emphasis on AI-tools utilized as a supplement to ones practice, and clear scopes around confidentiality and guidance. The study recommends 5 main considerations towards policies pertaining to the use of AI-tools in clinical settings with children and youth: (1) policy stance messaging emphasizes "AI-Assisted, clinician-led care", (2) confidentiality guidelines in using AI-tools, (3) tiering around identifying when tools can be used, (4) implementation of mandatory training around using AI-tools, and (5) organization-wise accountability measures to maintain the use of AI-tools. Overall, while AI-tools have benefits in clinical settings supporting child and youth mental health, policies and guidance may provide guidance and clarity for clinicians who choose to use AI-tools in their practice.