Public Responses to the First Randomized Controlled Trial of a Generative Artificial Intelligence Mental Health Chatbot

We conducted a qualitative analysis of comments posted on news articles and blogs covering the first randomized controlled trial of a generative AI therapy chatbot. Four publications included comments (New York Times [n=258]; Yahoo News [n=36]; Medscape [n=6]; Philly Voice [n=3], total N=300). Negative reactions to AI-provided chatbots (n=213) were more common than positive ones (n=142). Negative comments included: (1) societal concerns (e.g., technology cannot solve structural drivers of mental illness and inaccessibility of care), (2) treatment effectiveness concerns (e.g., the human therapeutic relationship is key), and (3) safety and regulatory concerns (e.g., concerns about data privacy and quality control). Positive comments included: (1) AI avoids some of the downsides of the current mental healthcare system (e.g., mental healthcare is degraded, inaccessible and expensive), and (2) potential upsides to AI (e.g., accessibility and cost, benefits of AI like a lack of judgment). These comments provide insight into broader perspectives on AI therapy chatbots.