Interview Highlights with Nicholas Jacobson on Therabot and Evergreen
Current Status of Therabot (2026):
Therabot, developed since 2019, is now a generative AI-based therapy tool with multiple interfaces. Testing has shown it to be safe and effective for clinical populations, specifically for anxiety, depression, and those at high risk for eating disorders. The focus is on responsible deployment with clinical oversight to broaden access.
Benefits and Risks of AI Chatbots:
Jacobson points out that the unpredictability of AI can be both a risk and a benefit. While there are valid concerns about AI’s potential harm, he emphasizes that traditional psychotherapy can also lead to negative outcomes. He suggests that AI might offer advantages over standard clinical care.
Differentiating Evergreen from Therabot:
Evergreen is targeted at a broader audience, focusing on mental well-being and personal development rather than clinical populations. Developed by students for students, it tailors its approach using available data to enhance interactions and wellness strategies.
Concerns Over AI Regulation:
Jacobson discusses opposition to New Hampshire Senate Bill 640, which would necessitate clinician review of every interaction with therapy chatbots. He argues that this approach mischaracterizes AI as a clinician and shifts liability improperly. He calls for more thoughtful regulation focused on the efficacy and safety of mental health AI.
Future of Tools at Dartmouth:
Jacobson encourages Dartmouth students to engage with the upcoming trials for Evergreen, which will explore both structured and generative interfaces. Ensuring safety and effectiveness is a priority as they prepare these tools for broader use.
Note: This summary highlights key points from an interview with Nicholas Jacobson on the development and implications of AI mental health tools.
