Don’t Miss the 2nd SODP Symposium on Friday, December 12
This year, the theme is AI, and the lineup includes six speakers. Each brings a different perspective on mental health and AI, ranging from research to lived experience, from a low- and middle-income countries perspective to public policy, and from personal benefits to safety concerns. We will be sharing links and research updates throughout the symposium too, and fielding your questions. Be sure to join us on December 12th at 1-3:30pm EST. Register here.
SODP Digital Navigator Training Update
The SODP team is working to finalize a standardized training for Digital Navigators. The training will cover a baseline set of skills and competencies for the role, with local teams able to add on additional training to customize the role. SODP will maintain a roster of trained and certified Digital Navigators and a library of additional teaching modules to help ensure the role expands and is well recognized.
A fascinating October 2025 paper explored the role of Digital Navigators in helping implement an app for people with schizophrenia. The research highlighted that changes in Digital Navigator support offered to patients do directly impact the ways an app can be effective for patients. The implications are clear, without standardized Digital Navigators it will be hard to know how effective many apps actually are. Read more about this study: Ben-Zeev D, Tauscher J, Sandel-Fernandez D, Buck B, Kopelovich S, Lyon AR, Chwastiak L, Marcus SC. Implementing mHealth for Schizophrenia in Community Mental Health Settings: Hybrid Type 3 Effectiveness-Implementation Trial. Psychiatric Services. 2025 Oct 24:appi-ps.
Have Your Say in AI and Mental Health Standards in 2026
Members should be on the lookout for a new survey in early 2026 that explores various proposals for AI standards. Those taking the survey and agreeing will be offered the chance to contribute to a resulting paper, or acknowledged if preferred. Not a SODP member yet? It's easy! Sign up here.
📈 Trends in Digital Mental Health
Perhaps regulation will replace the Taco Bell Test for AI
If you have not heard of the Taco Bell Test for AI, don’t worry, and see our last newsletter or the blog post to catch up. But perhaps we may not need such tests to assess AI anymore, as in the last months, regulators have been making their presence known in the AI mental health space. On November 6th, the US Food and Drug Administration held a daylong meeting of its Digital Health Advisory Committee on the topic of mental health AI. The recording can be found here.
Then on November 16th, there was a US Congressional oversight committee on AI and Mental Health: The recording can be found here. Also in late November, the UK Medicines and Healthcare products Regulatory Agency spoke at the eMental Health International Collaborative (eMHIC) meeting in Toronto and shared updates and guidance on digital mental health technology. Various US states, like California, shared their own policies seeking to ban AI chatbots for youth and there are certainly more related actions happening across the world.
While many of these regulatory actions are early steps, they are notable, as just one year ago there were few regulatory actions. As these efforts crystallize in 2026, we likely will not need the Taco Bell test to help judge when AI is ready for mental health. But until then, the SODP community should keep one eye on the regulatory space and the other on the drive-through.
⁉️ Hot Question in Digital Mental Health
This Edition's Question
Question: Is an LLM-based AI chatbot better than a rule-based AI chatbot?
Answer: No. While the initial generation of rule-based chatbots has quickly vanished and now LLM chatbots are popular, this does not make LLMs inherently safer or more effective for mental health. Rule-based chatbots have some advantages, including that they are less likely to go ‘off script’ and cause harm. And for some cases in mental health, staying on script may be okay, like offering a manualized therapy, perhaps. While LLM chatbots can, of course, do more, sometimes more can become too much, as we have seen in the media with cases covered of AI psychosis and even lawsuits alleging deaths by suicide. There is much we still don’t know about LLMs and that also makes them risky. Of course, there is a middle ground with machine learning-based chatbots, which offer some advantages of the other two types but also have some limitations. To learn more about evaluating rule-based, machine learning, and AI chatbots, read more here.
Do you have a question you want answered in the next newsletter? Let us know here.
✈️ Upcoming Events
☝️ See above for the December 12 SODP Symposium on AI and register here.
The 2026 American Psychiatric Association (APA) Annual Meeting will be held May 16-20, 2026 in San Francisco, California. This is one of the largest and most established psychiatry conferences, drawing clinicians, researchers, policymakers, and industry leaders. The meeting features a growing programming in digital psychiatry, including sessions on mobile mental-health tools, digital phenotyping, AI in clinical care, telepsychiatry, and more.
🌐 Upcoming Webinar Spotlight
Join JMIR Mental Health and the Society of Digital Psychiatry on Thursday, January 29 at 11 am EST for a live virtual webinar with Dr. Brian E. Bunnell, PhD, on enhancing engagement with digital mental health tools.
Get your brand in front of highly engaged clinicians—consider advertising with SODP. In addition to our membership journal newsletter, we offer a society website and access to more than 32,000 online professionals. Email dennis.obrien@jmir.org to learn more about running your first campaign with us.
-1.png?upscale=true&width=1200&upscale=true&name=Untitled%20(1000%20x%20400%20px)-1.png)
