JAMA + AI Weekly Update January 30, 2026

Dear readers,

As JAMA+ AI’s Editor in Chief, I mostly focus on curating the best artificial intelligence content from across the JAMA Network. What readers may not know is that I am a psychiatrist by training, with more than 25 years in clinical practice. As millions turn to AI for mental health treatment, I have followed the debates among my colleagues and realized that this sounds awfully familiar.

To understand the history of chatbots in mental health, I went straight to the source. In this week’s episode of JAMA+ AI Conversations, you’ll hear my interview with a modern chatbot, conducted in November, about some of the earliest efforts to make chatbot therapists.

Thanks for subscribing,
Roy Perlis, MD, MSc

Hear JAMA+ AI Editor in Chief’s interview with a chatbot

Editor’s Picks in this week’s JAMA+ AI:   

• An analysis of over 100,000 submissions to 13 JAMA Network journals revealed AI usage disclosures by authors rose from 1.7% to nearly 6% over 27 months, most commonly for language refinement. Manuscripts from non–English-speaking countries, those ultimately withdrawn or rejected, and certain article types were more likely to disclose AI involvement.
• A similar analysis of manuscripts submitted to 49 BMJ journals between April and November 2024 found that 5.7% of submissions included authors-declared use of AI. Chatbots were most frequent, with writing assistants also prominent.
• In a JAMA Editor’s Note highlighting findings from both Research Letters, Deputy Editors Preeti Malani, MD, MSJ, and Joseph Ross, MD, MHS, write that journal editors and publishers should anticipate increases in author AI use and work to develop guidance to promote transparency and preserve research integrity.