JMIR Mental Health

As digital communication becomes central to daily life, psychotherapy increasingly has access to patients’ electronic media data. Although digital phenotyping has been widely studied, less is known about whether incorporating personal communication data, such as text messages, improves clinical outcomes in psychotherapy.

Design research on AI technologies for mental health has expanded rapidly, yet existing work has largely focused on practitioners’ perspectives within isolated therapeutic stages or specific services. As a result, there is limited understanding of how practitioners view the role of AI across the full therapeutic journey, including assessment, diagnosis, and treatment, or how these perspectives can be translated into actionable and ethically grounded design guidance. This lack of stage-specific guidance introduces several risks: (1) the development of AI tools that practitioners may not trust or adopt, (2) heightened ethical concerns, particularly in diagnosis where bias and erosion of human judgment are most likely, and (3) missed opportunities to leverage AI to support ongoing assessment and treatment adaptation.

Decentralized (ie, fully remote) clinical trials (DCTs) offer a scalable and ecologically valid approach to evaluating digital mental health interventions (DMHIs). Yet, incorporating objective measures of well-being within DCTs poses novel methodological and feasibility challenges that can limit the rigor and precision of data collection in the field.

Mental illness contributes substantially to global disability, and public adoption of AI for mental health support is accelerating without commensurate safety evaluation. General-purpose large language models can fail to recognize naturalistic text expressing suicidal ideation, with immediate clinical consequences. Domain-specific natural language processing methods offer a contrasting approach, but ensembles integrating the two have not been formally characterized.

Mental health disorders affect 970 million people globally; yet, over 50% do not access timely evaluation due to structural barriers and professional shortages. Chatbots and AI-based conversational agents have emerged as promising tools for mental health screening and assessment.

Obsessive-compulsive disorder (OCD) is a common psychiatric disorder, with two-thirds of affected individuals reporting severe impairment. Despite its substantial burden and moderate heritability, the etiology of OCD remains poorly understood, and treatments are often suboptimal. Although recent genome-wide association studies (GWAS) have identified some risk loci, much of the genetic architecture of OCD remains undiscovered, underscoring the need for scalable approaches to identify large, well-defined patient cohorts.

Generative AI (GenAI) tools, such as OpenAI’s ChatGPT, Google’s Gemini, and Microsoft Copilot, are being adopted across health care settings. Despite the popularity of GenAI, empirical data on the extent of adoption by mental health clinicians in routine psychotherapy practice are scarce.

Perinatal mental health disorders affect approximately 20% of pregnant and postpartum individuals, and are associated with substantial maternal and infant morbidity. Traditional assessment relies on infrequent, subjective self-reports. Mobile devices, including smartphones and wearables, offer opportunities for continuous and objective measurement, but evidence on their assessment utility in perinatal populations remains fragmented.

Borderline personality disorder (BPD) is associated with substantial distress and a high risk for suicide. Individuals with BPD may be unable to access evidence-based treatments like dialectical behavior therapy (DBT). Artificial intelligence conversational agents (AI-CA) are increasingly discussed as scalable tools for mental health support, but little is known about how DBT clinicians understand the possible role of AI-CA in treatment.

Machine learning and natural language processing have demonstrated significant potential for mental health assessment: describing your mental health in your own words can offer a more ecologically valid approach than traditional rating scales. However, most models focus on specific diagnoses, conditions, or symptoms, which may prematurely assign labels and potentially reinforce stigma in the context of early-stage mental health screening.

Mental health clinical notes contain decision-critical information often absent from structured electronic health record fields. Large language models (LLMs) can extract clinically relevant signals from narrative text; however, variability in output format, limited reproducibility, and inconsistent evaluation remain barriers to clinical deployment. Despite rapid advances in LLM-based information extraction, clear and reproducible guidance for interdisciplinary clinical teams is limited.

Self-harm, including suicidal thoughts, self-injurious behavior, and disordered eating, is a major public health concern among adolescents and young adults. Youth increasingly encounter self-harm–related content online, including websites that explicitly encourage harmful behaviors. Although exposure to such content has been linked to poorer mental health, most research is cross-sectional and has not examined longitudinal exposure trajectories. Whether persistent exposure represents a distinct digital risk pattern remains unclear.