GenAI Use Increases Companion Interactions Among High-Risk

A 2026 cross-sectional survey of 1,003 US young adults by Buck et al. found that individuals with elevated psychosis risk were not more likely to have ever used generative AI, but reported significantly more intensive use and frequent chatbot conversations. Elevated-risk respondents more often sought social or emotional support, assigned human-like roles to chatbots, and reported delusion-related interactions (13.3%–30.7%), suggesting chatbots may affect symptom experiences.