
At a Glance
- Therabot significantly reduces symptoms of depression, anxiety, and eating disorders
- Dartmouth College conducted the first Randomized Controlled Trial (RCT) proving its efficacy
- Therabot delivers Cognitive Behavioral Therapy (CBT) via a text-based app
- Participants experienced notable symptom reduction after eight weeks
AI Therapy Breakthrough
In a groundbreaking study, Dartmouth College researchers have unveiled Therabot, an AI-powered chatbot designed to deliver Cognitive Behavioral Therapy (CBT) through a text-based application. The randomized controlled trial, published in NEJM AI, demonstrated that Therabot significantly reduces symptoms of major depressive disorder, generalized anxiety disorder, and eating disorders over an eight-week period.
The study involved 210 adults with clinically significant symptoms, divided into a control group and a group that interacted with Therabot. Participants using Therabot reported a 51% reduction in depression symptoms, a 31% decrease in anxiety, and a 19% improvement in eating disorder concerns, according to Dartmouth’s official report.
Watch Dartmouth’s coverage on the clinical trial at First Therapy Chatbot Trial Yields Mental Health Benefits.
Personalized Support at Scale
Therabot operates as a mobile application available on iOS and Android platforms. It utilizes Large Language Models (LLMs) to generate human-like interactions, enabling conversations based on CBT principles. As described in Psychology Today, the chatbot aims to replicate traditional therapy experiences and expand access to underserved populations.
Participants engaged with Therabot for an average of six hours over the trial period—roughly the equivalent of eight therapy sessions. Users reported forming a “therapeutic alliance” with the AI, comparable to the relationship typically developed with a human therapist, according to Dartmouth researchers.
Addressing the Mental Health Provider Shortage
The emergence of Therabot responds to a national crisis: fewer than one-third of Americans have adequate access to mental health care. Dartmouth engineers developed Therabot to bridge this gap after earlier AI therapy efforts fell short. As noted in Digital Health Insights, Therabot’s design overcomes typical limitations in digital therapeutics like poor retention and low user engagement.
Dartmouth senior researcher Nicholas Jacobson and his team emphasized in a Geisel School press release that AI chatbots hold “promise for building highly personalized, effective mental health treatments at scale.”
Though Therabot shows strong results, the study maintained human oversight to ensure safety—an approach researchers say remains essential. As detailed in DHInsights, future trials will need broader demographic samples and potentially include autonomous options.
Therabot marks a pivotal step in transforming mental health support, opening doors to scalable, AI-assisted therapy for millions in need.