Artificial intelligence use during suicidal crises in adolescents and young adults: a case series from emergency psychiatry

Key Findings

• — Seven patients aged 15 to 23 years (six female) were identified; diagnoses were major depressive disorder (n = 3), post-traumatic stress disorder (n = 2), adjustment disorder (n = 1) and unspecified emotional disorder (n = 1).
• — Six of seven interactions involved seeking information about the lethality of medications or other substances.
• — Based on patients' accounts, chatbot responses were classified as potentially harmful (n = 5), ambiguous or incomplete (n = 1) or protective (n = 1).
• — The temporal relationship between chatbot interactions and subsequent clinical events varied across cases, and the authors state that influence on subsequent behaviour cannot be established.

Methodology Notes

Single-site retrospective case series, n = 7, cases identified from medical records documenting chatbot interactions during routine clinical assessment; classification rests on patients' accounts, not on chat logs. Written consent from all seven patients and parental consent for the three minors; no ethics committee approval required under French regulation for case series (per the article). Verified through the Springer-hosted full text (r.jina text of the BMC page, 9,187 bytes) and the Crossref abstract; the BMC page itself returned a bot-challenge page to curl. Curator re-read the article text on 2026-11-10: the BMC page is an early accepted-article version that will be replaced by the version of record.

Sources