Answers are grounded in this paper and cite their sources.
Explores the impact of general-purpose LLMs on mental health outcomes, suggesting urgent oversight is needed.
General-purpose large language models (LLMs) like ChatGPT were never intended to be therapists; they were built as general-purpose assistants. Yet when human help is not available, many people now turn to these tools for advice, comfort, and coping. This shift is happening without oversight.1 Experts have warned that general-purpose LLMs are becoming de facto mental health aids without evidence-based guardrails.2 So far, psychiatry’s response has been slow and fragmented, a blind spot that needs to be addressed urgently.
No takes yet. Share an insight, caveat, or question.
Vinh et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: