John Torous
Verified
Psychiatrist, Director BIDMC DigitalPsych.org, Associate Professor @HarvardMed, Editor JMIR MentalHealth, Alum @Berkeley_EECS
Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by John Torous
A scoping review on the mental health harms of LLM-based chatbots
Abstract Chatbots based on large language models (LLMs) are increasingly used in everyday life. Although concerns about negative impacts on mental health are raised, a synthesis of the research and themes of the growing field of potential mental health harms of LLM-based chatbots has not yet been conducted.
LLMs as Clinical Instruments—Toward Verifiable Reasoning
Large language models (LLMs) have moved from demonstration to deployment in psychiatric settings. They are now used to draft notes, summarize longitudinal records, generate patient education, support inbox triage, and synthesize literature. The question is no longer whether mental health clinicians will encounter LLMs, but whether they can use rapidly improving systems in ways that improve care without amplifying hidden error.
Public availability of evidence-based mental health apps: a systematic review and meta-analysis of self-guided apps
Abstract Despite mental health apps having a rigorous evidence base, little is known about the proportion of efficacious apps that are publicly available. Our systematic review and meta-analysis identified self-guided mental health apps evaluated in controlled trials, determined their public availability, and examined how evidence quality and strength influence availability. A literature search identified 16,971 records, of which 110 studies, investigating 112 apps, met inclusion criteria.
AI-Associated Psychotic Phenomena: Collect Before You Classify
No abstract is available for this article. Data Availability Statement Data sharing not applicable to this article as no datasets were generated or analyzed during the current study. References 1, “Generative Artificial Intelligence Chatbots and Delusions: From Guesswork to Emerging Cases,” Acta Psychiatrica Scandinavica 152, no. 4 (2025): 257–259, https://doi.org/10.1111/acps.70022.
Barriers to understanding how many people use AI for mental health support: an estimate and narrative review
Abstract Reports suggest 3–70% of AI users utilize AI for mental health. A more precise estimate is needed to guide public health policy, research, and to understand the limitations of producing an estimate. We approximate that 27% of AI users utilize AI for mental health support. However, we acknowledge that heterogeneous data sources and a lack of nosology standardization confound estimates. Consensus building will be key before better prevalence estimates are possible.
Specialized or General-Purpose—The Wrong Question for Mental Health AI Safety
That artificial intelligence (AI) can cause harm in mental health contexts is no longer contested.1 The debate has shifted to what kinds of harm, how to measure them, and who is accountable. Early AI models and their associated guardrails were largely inadequate at addressing mental health–related harm, but today that is beginning to change. Advances in both model capability and the surrounding safety infrastructure now permit a more proactive approach.1
Lessons learned from visiting an AI-powered store
A broken mug sits on a desk in the Division of Digital Psychiatry at Beth Israel Deaconess Medical Center. Printed across the surviving ceramic shard are the words: "This was made by AI." Our team thinks that the story of this mug neatly captures the hype, hope, and even harm surrounding artificial intelligence. At the Division of Digital Psychiatry, we spend our days studying how digital technologies are shaping mental healthcare.
A New CMS Payment Model for AI-Delivered Behavioral Health Care
The mismatch between behavioral health demand and supply in the US is no longer news, and the field has spent decades exploring ways to expand access to care. What is news is who is showing up to fill the gap. Consumer-facing behavioral health technology companies and general-purpose generative artificial intelligence (AI) systems have become a common front door for behavioral health support, sometimes by design and often inadvertently.
Accelerating Digital Mental Health: The Society of Digital Psychiatry's Three-Pronged Road Map for Education, Digital Navigators, and AI
Faculty, Staff and Student Publications Publication Date 11-27-2025 Journal JMIR Mental Health PubMedCentral® Posted Date 11-27-2025 PubMedCentral® Full Text Version Post-print Abstract Digital mental health tools such as apps, virtual reality, and artificial intelligence (AI) hold great promise but continue to face barriers to widespread clinical adoption.
Transforming artificial intelligence into artificial wisdom
|
Abstract A rapidly expanding behavioral epidemic of loneliness is profoundly affecting mental and physical health of individuals and communities worldwide. Empirical research shows that, unlike intelligence, human wisdom can mitigate loneliness and promote mental well-being. However, there is a severe global shortage of mental healthcare workforce required to support population-level well-being.
By Dilip V. Jeste, Martin P. Paulus Verified, George S. Alexopoulos, Willem M Otte, Yuto Satake, Peter Jongho Na, John Torous Verified, Ante Prodan, Jo-An Occhipinti, Marcel Barnard
|
Nature
Verified
Integrating Data Visualizations Into Digital Mental Health Care for Adults With Anxiety and Depression: Participatory Design and Case Study
Ethical Considerations Participants and Clinical Context Figure 1. Digital clinic care model workflow.
The Society of Digital Psychiatry AI Psychosis Survey Results
Being part of the Society of Digital Psychiatry means you help shape and lead important conversations in mental health. A recent survey, conducted in collaboration with JMIR Publications, collected members' insights and perspectives on the emerging phenomenon of “AI Psychosis”. The survey collected 97 submissions in January and February 2026. A summary of these perspectives will be included in a future News and Perspectives article in the Journal of Medical Internet Research.
A scoping review of the characteristics, responsibilities, implementations and evaluations of digital navigators in healthcare
Abstract The emergence of various tools and technologies has contributed to the creation of new roles in healthcare, such as digital navigators, who support both clinicians and patients with the adoption and sustained use of these technologies. The digital navigator role is gaining traction for its potential to be adapted and scaled across diverse healthcare settings. However, the formalization of digital navigator roles in healthcare remains limited.
An AI-based mental health guardrail and dataset for identifying psychiatric crises in text-based conversations
Abstract Large language models often mishandle psychiatric emergencies, offering harmful or inappropriate advice. This study evaluated the Verily Mental Health Guardrail (VMHG) on two clinician-labeled datasets: the Verily Mental Health Crisis Dataset v1.0, containing 1800 simulated messages and the NVIDIA Aegis AI Content Safety Dataset subsetted to 794 mental health-related messages. Performance was benchmarked against OpenAI Omni Moderation Latest and NVIDIA NeMo Guardrails.
Beyond artificial intelligence psychosis: a functional typology of large language model-associated psychotic phenomena
Introduction Although the term artificial intelligence (AI) psychosis has become widespread in media reporting on large language model (LLM)-associated psychotic phenomena, it represents neither a clinical diagnosis nor a unified phenomenon. Major news outlets have documented several cases including LLM-associated suicides,1,2 sudden-onset delusions in previously healthy individuals,3,4 and LLM-facilitated violence,5,6 often grouping these disparate events under the single label of AI psychosis.
LINC: a framework for maintaining high-quality passive data in digital phenotyping studies - Scientific Reports
Abstract Achieving high-quality passive data in smartphone-based digital phenotyping studies remains a significant challenge in real-world settings where technical barriers and inconsistent user engagement limit continuous data collection. In response to the field’s overreliance on imputation methods and model-based approaches, we present LINC, a framework of best practices that supports the collection of high-quality passive data in digital phenotyping research.
Do generative AI chatbots increase psychosis risk?
A growing body of clinical trials and meta-analyses shows that generative artificial intelligence (AI) chatbots can reduce psychiatric symptoms, confirming their therapeutic potential. The vast majority of this research is case based, with low quality evidence. More clinically focused research has demonstrated reductions in anxiety and depression scores, but only against a waitlist control group.
Four reasons why generative AI chatbots could lead to psychosis in vulnerable people
Now That You Have iOS 26, Go Change These 10 iPhone Settings Change these settings and you'll feel like you bought a new iPhone. Now that the iPhone's latest mobile operating system, iOS 26, is officially out, you can finally try its new Liquid Glass interface and advanced AI features. But before you dive into all the new toys, it's the perfect time to …
Large language models as mental health providers
Get full text access Log in, subscribe or purchase for full access. References 1. Rousmaniere, T ∙ Zhang, Y ∙ Li, X ∙ et al. Large language models as mental health resources: patterns of use in the United States Practice Innovations. 2025; 2. Zao-Sanders, M How people are really using Gen AI. Harvard Business Review 3. Hua, Y ∙ Na, H ∙ Li, Z ∙ et al. A scoping review of large language models for generative tasks in mental health care NPJ Digit Med. 2025; 8:230 4. Jin, Y ∙ Liu, J ∙ Li, P ∙ et al.
Bridging the Digital Divide: Integrating Apps into Effective Mental Health Care
How can we blend traditional telehealth sessions with smartphone applications while addressing digital literacy barriers to create more accessible and effective mental health treatment for all patients? For permission requests, please contact NEJM Reprints at reprints@nejm.org Notes John Torous has nothing to disclose.
Assessing generative artificial intelligence for mental health
FREE with registration! Log in or register to access the full article. Further reading Goldberg et al., 2023. Goldberg, SB ∙ Sun, S ∙ Carlbring, P ∙ et al. Selecting and describing control conditions in mobile health randomized controlled trials: a proposed typology NPJ Digital Med. 2023; 6:181 Heinz et al., 2025. Heinz, MV ∙ Mackin, DM ∙ Trudeau, BM ∙ et al. Randomized trial of a generative AI chatbot for mental health treatment NEJM AI. 2025; Na et al., 2025. Na, H ∙ Hua, Y ∙ Wang, Z ∙ et al.
Integrating Artificial Intelligence and Smartphone Technology to Enhance Personalized Assessment and Treatment for Eating Disorders
Smartphones can augment assessment, detection, and treatment for eating disorders, yet this technology remains underutilized in this field and has yet to reach its potential. Sophisticated technological capabilities like digital phenotyping and natural language processing may aid the detection of eating disorders in new ways while also informing the development of closed-loop, adaptive interventions.
Digital measures of activity and motivation impact depression and anxiety in the real world - npj Digital Medicine
Abstract Mood and anxiety disorders are highly comorbid, with symptom severity varying over time. Individuals with and without these disorders completed 30-days of ecological momentary assessment (EMAs) of depression, anxiety and distress, developed based on the established Mood and Anxiety Symptom Questionnaire (MASQ).
Enabling FAIR data stewardship in complex international multi-site studies: Data Operations for the Accelerating Medicines Partnership® Schizophrenia Program | Schizophrenia
|
Abstract Modern research management, particularly for publicly funded studies, assumes a data governance model in which grantees are considered stewards rather than owners of important data sets. Thus, there is an expectation that collected data are shared as widely as possible with the general research community. This presents problems in complex studies that involve sensitive health information. The latter requires balancing participant privacy with the needs of the research community.
By Grace R. Jacobs, Nora Penzel, Dominic B. Dwyer, Hok Pan Yuen, Cheryl Corcoran, Omar John, Jean Addington, Kelly Allott, Scott Clark, Michael P. Harms, Diana O Perkins, John Torous Verified, Alison R. Yung, Vijay A. Mittal, Jai Shah, Daniel Wolf, Guillermo Cecchi, Marek R. Kubicki, Kathryn Eve Lewandowski, Carrie E. Bearden, René S. Kahn, Scott Woods, Martha E. Shenton, Dheshan Mohandass
|
Nature
Verified
Data analysis strategies for the Accelerating Medicines Partnership® Schizophrenia Program - Schizophrenia
|
Abstract The Accelerating Medicines Partnership® Schizophrenia (AMP® SCZ) project assesses a large sample of individuals at clinical high-risk for developing psychosis (CHR) and community controls. Subjects are enrolled in 43 sites across 5 continents. The assessments include domains similar to those acquired in previous CHR studies along with novel domains that are collected longitudinally across a period of 2 years.
By Nora Penzel, Jean Addington, Celso Arango, Sylvain Bouix, Dylan Campbell, Eduardo Castro, Cheryl Corcoran, Dominic B. Dwyer, Sophia Frangou, Michael P. Harms, Grace R. Jacobs, Joseph Kambeitz, Nikolaos Koutsouleris, Jun Soo Kwon, Kathryn E. Lewandowski, Vijay A. Mittal, Diana O Perkins, Johanna Seitz-Holland, John Torous Verified, Jijun Wang, Daniel Wolf, Carrie E. Bearden, Scott Woods, René S. Kahn, Martha E. Shenton, Guillermo Cecchi
|
Nature
Verified
From statistics to deep learning: Using large language models in psychiatric research
1 INTRODUCTION The integration of Large Language Models (LLMs) into psychiatry and mental health research is poised to revolutionize the field by enhancing diagnostic accuracy, personalizing care, and streamlining administrative tasks (Hua et al., 2024; Obradovich et al., 2024; Zhou et al., 2023). LLMs can efficiently process vast amounts of data, summarize clinical notes, and assist in complex decision-making processes (Yu et al., 2023).
Texting abbreviations perceived as insincere, associated with lower response rates
You've successfully added Practice Management in Psychiatry to your alerts. You will receive an email when new content is published. Click Here to Manage Email Alerts We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com. Key takeaways: The use of texting abbreviations was associated with lower perceptions of effort.
Special Report: Understanding the Architecture of AI
Full access Clinical & Research To see how AI might fit into psychiatric practice, it is important to first understand the entire ecosystem—comprising data, technologies, and organizational decisions—that underpins AI system development. Data: Foundation models are trained on large datasets (some estimated at trillions of tokens), which are often sourced from the internet and may be inherently biased.
Special Report: Are You Ready for Generative AI in Psychiatric Practice?
Full access Clinical & Research The advances and promises around AI have skyrocketed, with billions of dollars already invested and much more expected. OpenAI’s ChatGPT has become a household name, extending its influence into various industries and professions, including psychiatry. A recent survey of 138 psychiatrists affiliated with APA revealed high use and expectation for ChatGPT in clinical question-answering, administrative tasks, and training.
Hybrid care in mental health: a framework for understanding care, research, and future opportunities - NPP—Digital Psychiatry and Neuroscience Original
Abstract Technology is playing an increasing role in healthcare, especially in mental health. Traditional mental healthcare, whether in-person or via telehealth, cannot by itself address the massive need for services. Standalone technology such as smartphone apps, while easily accessible, have seen limited engagement and efficacy on their own.
Large language models outperform mental and medical health care professionals in identifying obsessive-compulsive disorder - npj Digital Medicine Original
Abstract Despite the promising capacity of large language model (LLM)-powered chatbots to diagnose diseases, they have not been tested for obsessive-compulsive disorder (OCD). We assessed the diagnostic accuracy of LLMs in OCD using vignettes and found that LLMs outperformed medical and mental health professionals. This highlights the potential benefit of LLMs in assisting in the timely and accurate diagnosis of OCD, which usually entails a long delay in diagnosis and treatment.
Social Isolation and Loneliness: Modern Pandemic of a Psychosocial Determinant of Health
Information Copyright 2024, SLACK Incorporated Address correspondence to Mei-Hua Hall, PhD, Department of Psychiatry, Harvard Medical School, McLean Hospital, 115 Mill Street, Belmont, MA 02478; email: mhall@mclean.harvard.edu.
Evidencia actual sobre la eficacia de las aplicaciones de salud mental para teléfonos inteligentes para los síntomas de depresión y ansiedad
La atención de salud mental para la depresión y la ansiedad ha evolucionado significativamente con la revolución tecnológica, particularmente con el surgimiento de aplicaciones para teléfonos inteligentes como herramientas escalables. Un metanálisis reciente, que incorpora más de 100 nuevos ensayos controlados aleatorios, buscó brindar estimaciones más precisas de los efectos de estas aplicaciones y comprender mejor su impacto.
La inestable dicotomía de la salud mental digital: bienestar y salud
Me Gusta 0 0 Comentar 0 0 Autor/autores: John Torous, Joseph Firth, Simon B. Goldberg...(et.al) Artículo | 20/05/2024 En la última década, las tecnologías digitales de salud mental (DMHT) como aplicaciones móviles, realidad virtual y dispositivos portátiles, junto con la inteligencia artificial generativa, han expandido su alcance. Estas herramientas buscan ofrecer intervenciones de salud mental escalables y mejorar el acceso a la atención. Sin embargo, su eficacia cl&ia...
Interindividual Sleep Variability Across the Psychosis Spectrum : A Systematic Review and Meta-Analysis
|
10. Aronica R, Ostinelli EG, Austin C, et al. Digital sleep phenotype and wrist actigraphy in individuals at clinical high risk for psychosis and people with schizophrenia spectrum disorders: a systematic review and meta-analysis. BMJ Ment Health. 2025;28(1):e301337. doi:10.1136/bmjment-2024-301337 PubMedGoogle ScholarCrossref 12. Sasidharan A, Kumar S, Nair AK, et al.
By Rosario Aronica, John Torous Verified, Amedeo Minichino, Melinda Mills, Philip M. McGuire, Dominic Oliver
|
JAMA Network
Verified
Digital Mental Health’s Unstable Dichotomy
Digital Mental Health’s Unstable Dichotomy—Wellness and Health For at least the last decade, digital mental health technologies (DMHTs) such as smartphone apps, virtual reality, and wearables have been expanding in scope and potential. Today, generative artificial intelligence (AI) has joined the list of guided (coached) and unguided (self-help) digital tools that aim to deliver mental health interventions.
Patient and Clinician Experiences with Sharing Data Visualizations Integrated into Mental Health Treatment
All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess.
Integrated Digital Mental Health
Back to Journals » International Journal of General Medicine » Volume 17 Perspectives General Medicine Authors Lim CT, Fuchs C, Torous J Received 12 November 2023 Accepted for publication 10 January 2024 Published 1 February 2024 Volume 2024:17 Pages 359—365 DOI https://doi.org/10.2147/IJGM.S449474 Checked for plagiarism Yes Review by Single anonymous peer review Peer reviewer comments 2 Editor who approved publication: Dr Woon-Man Kung Christopher T Lim,1,2 Cara Fuchs,1 John Torous3...
ChatGPT and mental healthcare: balancing benefits with risks of harms
ChatGPT and mental healthcare: balancing benefits with risks of harms The WHO estimates that worldwide one in eight people live with mental illness.1 Stigmatisation and human rights violations, combined with lack of resources, including shortfalls in mental health professionals, pose significant barriers to psychiatric care.2 Consequently, clinician time is one of the scarcest resources in healthcare, and psychiatrists report looking to advances in artificial intelligence (AI) to improve...
Correction: Digital and precision clinical trials: innovations for testing mental health medications, devices, and psychosocial treatments - Neuropsychopharmacology
Correction to: Neuropsychopharmacology https://doi.org/10.1038/s41386-023-01664-7, published online 07 August 2023 In Figure 1, the text boxes on the left inadvertently repeated the same information. The figure has now been replaced with an updated version. The original article has been corrected. About this article Lenze, E., Torous, J. & Arean, P. Correction: Digital and precision clinical trials: innovations for testing mental health medications, devices, and psychosocial treatments.
The Digital Clinic: An Innovative Mental Health Care Delivery Model Utilizing Hybrid Synchronous and Asynchronous Treatment
Published August 16, 2023 NEJM Catal Innov Care Deliv 2023;4(9) DOI: 10.1056/CAT.23.0100 Notes Natalia Macrynikola, Nicholas Nguyen, Erlend Lane, and Shirley Yen have nothing to disclose. John Torous is a scientific advisor for Precision Mental Wellness, Cambridge, Massachusetts (not discussed in this article).
Digital and precision clinical trials: innovations for testing mental health medications, devices, and psychosocial treatments - Neuropsychopharmacology
Author notes These authors contributed equally: John Torous, Patricia Arean.
An exploratory analysis of the effect size of the mobile mental health Application, mindLAMP
Skip to main page content Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation Save citation to file Format: Add to My Bibliography Your saved search Name of saved search: Search terms: Frequency: Which day? Which day?
Increasing the value of digital phenotyping through reducing missingness: a retrospective review and analysis of prior studies
Skip to main page content Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation Save citation to file Format: Add to My Bibliography Your saved search Name of saved search: Search terms: Frequency: Which day? Which day? Report format: Send at most: Send even when there aren't any new results Optional text in email: Create a file for external citation management software doi: 10.1136/bmjment-2023-300718. Affiliations Affiliations 1 Harvard Medical School, Boston, Massachusetts, USA.
Increasing the value of digital phenotyping through reducing missingness: a retrospective review and analysis of prior studies
Suggestions Digital phenotyping research requires careful consideration to maximise data coverage. Solutions include recruiting additional participants12 to increase the chance of higher data coverage and study protocols designed to maximise data coverage from enrolled participants. While imputation methods can also help address some issues related to missingness18 after data collection, solutions to maximise data coverage are the focus of our recommendations.
This mental health app could change how we understand and treat schizophrenia
When I was training in psychiatry, it was always tricky seeing patients in the emergency department. We would try to understand their situation as quickly as we could, but you are often working to make the best decision with limited information. Then one day a patient came in and it sounded like something very bad had happened in a personal relationship. The patient agreed to let us read their text messages and immediately it was very clear what was happening.
Usable Data Visualization for Digital Biomarkers: An Analysis of Usability, Data Sharing, and Clinician Contact
Background The role of digital data and biomarkers for healthcare continues to expand. Smartphones and other connected devices are able to collect rich, real-time, and temporally dense data that have accelerated interest and potential in behavioral biomarkers. From simple longitudinal surveys (https://www.zotero.org/google-docs/?5jhI5l [1]) to interactive assessments of reaction time [2, 3], digital data from smartphones offer a new window into health.
Regulatory considerations to keep pace with innovation in digital health products - npj Digital Medicine
Abstract Rapid innovation and proliferation of software as a medical device have accelerated the clinical use of digital technologies across a wide array of medical conditions. Current regulatory pathways were developed for traditional (hardware) medical devices and offer a useful structure, but the evolution of digital devices requires concomitant innovation in regulatory approaches to maximize the potential benefits of these emerging technologies.
How to choose a mental health app
With the ease of access to healthcare via a smartphone, it is not surprising that more and more of us are turning to apps to get assistance with our mental health. Indeed, there are many more than 10,000 apps available to pick from today, from a variety of sources – including private companies and government agencies – and millions of people around the world are seeking to manage their mental health this way.
Get Up to Speed on Use of Technology in Psychiatric Care
As COVID-19 has and continues to drive so much care online and into digital formats, psychiatrists have had to quickly adapt to telehealth, smartphone apps, and other technologies. But now as it becomes possible to return to in-person care, psychiatrists have to decide what the future of the field will be and the next chapter of innovation in psychiatry.
Show More
loading
Actions
Get in touch with John
Contact John, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack