{"id":36279,"date":"2026-07-16T15:12:20","date_gmt":"2026-07-16T19:12:20","guid":{"rendered":"https:\/\/geiselmed.dartmouth.edu\/news\/?p=36279"},"modified":"2026-07-16T16:19:57","modified_gmt":"2026-07-16T20:19:57","slug":"nicholas-jacobson-recognized-for-foundational-therabot-trial","status":"publish","type":"post","link":"https:\/\/geiselmed.dartmouth.edu\/news\/2026\/nicholas-jacobson-recognized-for-foundational-therabot-trial\/","title":{"rendered":"Nicholas Jacobson Recognized for Foundational Therabot Trial"},"content":{"rendered":"
\"\"
Photo by Eli Burakian D'01<\/figcaption><\/figure>\n

Geisel professor Nicholas Jacobson<\/a>, in recognition of leading the first clinical trial<\/a> of a generative AI therapy chatbot using a platform developed in his lab called Therabot, has been named one of two finalists worldwide for the Chen Institute and Science Prize for AI Accelerated Research<\/a>.<\/p>\n

Jacobson, an associate professor of biomedical data science and psychiatry, will receive funding and publish a manuscript<\/a> in Science on July 16 about the significance of his work\u2014and on\u00a0 clinically developed AI-assisted therapy in general\u2014to expand access to mental health care.<\/p>\n

The prize, awarded by the Tianqiao and Chrissy Chen Institute in partnership with Science and the American Association for the Advancement of Science, recognizes early-career researchers working at the cutting edge of artificial intelligence and the natural sciences. The prize winner is Sergey Stavisky, an associate professor of neurological surgery at University of California Davis.<\/p>\n

\"There are a lot of folks rushing into AI for mental health, and it's easy to put out a proof-of-concept that looks great at first glance, but the safety and efficacy is not well established. We took the slower road with Therabot, grounding it in evidence-based practice and testing it in a randomized controlled trial, and it means a great deal that Science and the Chen Institute see the value in that approach,\" Jacobson says.<\/p>\n

\"There is no replacement for in-person care, but there are nowhere near enough providers to go around,\" he says. \"Recognition like this helps make the case that carefully built and carefully tested AI can reach the huge number of people who are outside the in-person care system.\"<\/p>\n

Jacobson is honored for his ongoing work on Therabot, which is a generative AI platform under development since 2019 in his AIM HIGH Lab<\/a> based in Geisel's Center for Technology and Behavioral Health<\/a>.<\/p>\n

In March 2025, Jacobson and his team published the first results of a randomized controlled trial that tests the efficacy and safety of AI-delivered psychotherapy.<\/p>\n

The trial enrolled 210 people, all of whom were diagnosed with major depressive disorder or generalized anxiety disorder, or were at clinically high risk for feeding and eating disorders. Of those, 106 participants were randomly selected to use Therabot for four weeks with unlimited access via a smartphone app. The researchers tracked a control group of 104 people who had no access to Therabot.<\/p>\n

People with Therabot access typed out responses to prompts about how they were feeling or initiated conversations. The platform responded with natural, open-ended dialog based on an original training set developed from current, evidence-based best practices for psychotherapy and cognitive behavioral therapy.<\/p>\n

The trial results, published in NEJM AI<\/a>, show that Therabot users experienced significant improvements in their symptoms. The results included an average 51% reduction in depressive symptoms, a 31% reduction in anxiety symptoms, and a 19% reduction in concerns related to eating disorders.<\/p>\n

But the researchers also caution that without clinician oversight, generative AI platforms can become dangerous to users experiencing a mental health crisis.<\/p>\n

Since the study's publication, Jacobson and his collaborator Michael Heinz<\/a>, an assistant professor of psychiatry and the paper's first author, have frequently spoken with legislators, regulators, and the media about improving the safety and efficacy of AI-assisted therapy by ensuring clinician oversight. The latest interview<\/a> with Jacobson can be heard on the July 16 episode of the Mental Health Evolution podcast<\/a>.<\/p>\n

Jacobson testified before the federal Food and Drug Administration's Digital Health Advisory Committee in November<\/a> (video<\/a>), and has also appeared before lawmakers<\/a> in New Hampshire.<\/p>\n

Therabot is currently in beta testing with multiple trials in different clinical settings either wrapping up or preparing to be launched. Jacobson and Heinz also co-founded Therabot Labs, LLC, to ultimately roll out the technology in accordance with clinical standards for monitoring safety and outcomes.<\/p>\n


\n