The First Trial of Generative AI Therapy Shows It Might Help With Depression

MIT Technology Review has published an in-depth analysis of the Dartmouth Therabot trial, placing it in the broader context of AI's potential role in addressing the global mental health crisis. The trial's results — a 51% reduction in depression symptoms and 31% reduction in anxiety symptoms — represent the strongest clinical evidence to date for generative AI as a therapeutic tool.

The review highlights what makes this trial significant compared to earlier chatbot studies: Therabot uses a modern generative AI model fine-tuned by clinical experts, delivering adaptive therapy rather than rigid scripted interactions. Participants reported trust levels comparable to working with human therapists, suggesting that the stigma barrier around AI mental health support may be lower than anticipated.

However, the analysis also raises important caveats. The four-week trial duration is short by clinical standards, and the study did not include patients with severe mental illness or active suicidality. The question of long-term efficacy — whether AI therapy produces durable improvements or requires ongoing engagement — remains unanswered.

The review contextualizes these results against the staggering scale of unmet mental health need: globally, an estimated 1 billion people live with a mental health condition, and the majority lack access to any professional support. Even imperfect AI tools that can provide evidence-based interventions at scale could have enormous public health impact.