“Feasible but Fragile”: An inflection point for artificial intelligence in mental health care.

Clegg, K. (2025).

Journal of Medical Internet Research, 27, e89202.

https://doi.org/10.2196/89202

On November 18, 2025, a congressional hearing was held in Washington, DC, by the US House Energy and Commerce’s Subcommittee on Oversight and Investigations, examining the risks and benefits of artificial intelligence (AI) chatbots.

Marlynn Wei, MD, JD; Jennifer King, PhD; and John Torous, MBI, MD (director of digital psychiatry at Beth Israel Deaconess Medical School and associate professor of psychiatry at Harvard Medical School) provided expert testimony at the congressional hearing. I sat down with Torous to discuss his reflections on the future of AI in mental health.

An Inflection Point

Following on the heels of several lawsuits and mounting concerns about the safety of commercially available AI chatbots and their widespread “off-label” use as psychological support, November’s congressional hearing was somewhat anomalous—in a good way.

“I actually am optimistic,” says Torous, “because we never saw a congressional oversight committee form in the early days when social media came out or when apps came out or when VR [virtual reality] came out. It’s exciting to see a body like Congress taking the time and attention to try to understand what the issue is.”

He remarks that it’s a different trajectory than we’ve seen over the past 25 years of digital health innovation, one that simultaneously signals that “we’re seeing the end of AI exceptionalism in mental health.” It suggests that regulators are taking the risks seriously and that AI—whether purpose-built or used de facto—will not be exempt from the same scrutiny applied to other clinical tools.

It’s also a potential inflection point from the otherwise rapid and underregulated growth and proliferation of AI tools for mental health, including many chatbots whose safety and efficacy remain to be definitively established. The shape of the trajectory now—whether these tools succeed or fail to materialize their potential for improving mental health care—depends on what we do from here.

The information is here.

Key Takeaways

• The future of artificial intelligence (AI) tools in mental health care is at an inflection point; regulators are taking both the potential benefits and the risks of these tools seriously.

• Whether these tools succeed or fail to meet their potential for improving mental health care depends on the extent to which stakeholders are able to successfully seize the moment and collaborate on transparent, high-quality research; establish and incentivize safety and efficacy; adopt patient-centric benchmarks; and think beyond traditional therapeutic models.

ChatGPT Clinical Use in Mental Health Care: Scoping Review of Empirical evidence.

Balan, R., & Gumpel, T. P. (2025).

JMIR Mental Health, 12, e81204.

https://doi.org/10.2196/81204

Abstract

Background:

As mental health challenges continue to rise globally, there is an increasing interest in the use of GPT models, such as ChatGPT, in mental health care. A few months after its release, tens of thousands of users interacted with GPT-based therapy bots, with mental health support identified as the primary use case. ChatGPT offers scalable and immediate support through natural language processing capabilities, but their clinical applicability, safety, and effectiveness remain underexplored.

Objective:

This scoping review aims to provide a comprehensive overview of the main clinical applications of ChatGPT in mental health care, along with the existing empirical evidence for its performance.

Methods:

A systematic search was conducted in 8 electronic databases in April 2025 to identify primary studies. Eligible studies included primary research, reporting on the evaluation of a ChatGPT clinical application implemented for a mental health care–specific purpose.

Results:

In total, 60 studies were included in this scoping review. The results highlighted that most applications used generic ChatGPT and focused on the detection of mental health problems and counseling and treatment. At the same time, only a minority of studies investigated ChatGPT use in clinical decision facilitation and prognosis tasks. Most of the studies were prompt experiments, in which standardized text inputs—designed to mimic clinical scenarios, patient descriptions, or practitioner queries—are submitted to ChatGPT to evaluate its performance in mental health-related tasks. In terms of performance, ChatGPT shows good accuracy in binary diagnostic classification and differential diagnosis, simulating therapeutic conversation, providing psychoeducation, and conducting specific therapeutic strategies. However, ChatGPT has significant limitations, particularly with more complex clinical presentations and its overly pessimistic prognostic outputs. Nevertheless, overall, when compared to mental health experts or other artificial intelligence models, ChatGPT approximates or surpasses their performance in conducting various clinical tasks. Finally, custom ChatGPT use was associated with better performance, especially in counseling and treatment tasks.

Conclusions:

While ChatGPT offers promising capabilities for mental health screening, psychoeducation, and structured therapeutic interactions, its current limitations highlight the need for caution in clinical adoption. These limitations also underscore the need for rigorous evaluation frameworks, model refinement, and safety protocols before broader clinical integration. Moreover, the variability in performance across versions, tasks, and diagnostic categories also invites a more nuanced reflection on the conditions under which ChatGPT can be safely and effectively integrated into mental health settings.

Jury finds Meta's platforms are harmful to children in 1st wave of social media addiction lawsuits

PBS News (2026, March 24).

SANTA FE, N.M. (AP) — A New Mexico jury found Tuesday that social media conglomerate Meta is harmful to children's mental health and in violation of state consumer protection law.

The landmark decision comes after a nearly seven-week trial. Jurors sided with state prosecutors who argued that Meta — which owns Instagram, Facebook and WhatsApp — prioritized profits over safety. The jury determined Meta violated parts of the state's Unfair Practices Act on accusations the company hid what it knew about about the dangers of child sexual exploitation on its platforms and impacts on child mental health.

The jury agreed with allegations that Meta made false or misleading statements and also agreed that Meta engaged in "unconscionable" trade practices that unfairly took advantage of the vulnerabilities of and inexperience of children.

Jurors found there were thousands of violations, each counting separately toward a penalty of $375 million.

Attorneys for Meta said the company discloses risks and makes efforts to weed out harmful content and experiences, while acknowledging that some bad material gets through its safety net.

The info is here.

Here are some thoughts:

A New Mexico jury ruled that Meta's platforms harmed children's mental health and violated state consumer protection law. After a seven-week trial, jurors found Meta prioritized profits over safety, made misleading statements, and exploited children's vulnerabilities — tallying thousands of violations worth $375 million in potential penalties. The verdict is part of a broader legal reckoning, with 40+ state attorneys general filing similar suits and a parallel federal case underway in California.

When corporations place profits above people, it's never the shareholders who pay the price. There have been multiple articles about Meta's harmful business practices.

Exploring the Ethical Challenges of Conversational AI in Mental Health Care: Scoping Review

Meadi, M. R., et al. (2025)

JMIR Mental Health, 12, e60432.

https://doi.org/10.2196/60432

Abstract

Background: Conversational artificial intelligence (CAI) is emerging as a promising digital technology for mental health care. CAI apps, such as psychotherapeutic chatbots, are available in app stores, but their use raises ethical concerns.

Objective: We aimed to provide a comprehensive overview of ethical considerations surrounding CAI as a therapist for individuals with mental health issues.

Methods: We conducted a systematic search across PubMed, Embase, APA PsycINFO, Web of Science, Scopus, the Philosopher’s Index, and ACM Digital Library databases. Our search comprised 3 elements: embodied artificial intelligence, ethics, and mental health. We defined CAI as a conversational agent that interacts with a person and uses artificial intelligence to formulate output. We included articles discussing the ethical challenges of CAI functioning in the role of a therapist for individuals with mental health issues. We added additional articles through snowball searching. We included articles in English or Dutch. All types of articles were considered except abstracts of symposia. Screening for eligibility was done by 2 independent researchers (MRM and TS or AvB). An initial charting form was created based on the expected considerations and revised and complemented during the charting process. The ethical challenges were divided into themes. When a concern occurred in more than 2 articles, we identified it as a distinct theme.

Conclusions: Our scoping review has comprehensively covered ethical aspects of CAI in mental health care. While certain themes remain underexplored and stakeholders’ perspectives are insufficiently represented, this study highlights critical areas for further research. These include evaluating the risks and benefits of CAI in comparison to human therapists, determining its appropriate roles in therapeutic contexts and its impact on care access, and addressing accountability. Addressing these gaps can inform normative analysis and guide the development of ethical guidelines for responsible CAI use in mental health care.

Here are some thoughts:

From a clinical perspective, the most immediate ethical tension identified in this review is the conflict between increasing accessibility and ensuring nonmaleficence (doing no harm). While proponents argue that CAI can bridge care gaps by offering constant availability and reaching those who fear stigma, the risks regarding safety and crisis management are profound. The review highlights that CAI systems often fail to contextualize user cues, leading to inappropriate responses in critical situations, such as suicidality. Furthermore, the phenomenon of AI "hallucinations"—where the system presents false information as fact—poses a unique danger in mental health, potentially exacerbating eating disorders or anxiety through misinformation. The lack of strong clinical evidence is also concerning; despite the commercial "hype," a significant portion of these tools have not been subjected to rigorous clinical studies to prove their efficacy compared to active controls.

Technologically, the "black box" problem creates a significant barrier to integrating CAI into professional practice. The review notes that the opacity of machine learning algorithms makes it difficult to explain how a CAI arrived at a specific therapeutic intervention, which undermines the principle of explicability and trust. This lack of transparency complicates accountability; if a CAI harms a patient, it remains unclear whether the responsibility lies with the developers, the deploying clinicians, or the algorithm itself—a concept known as the "responsibility gap". For board-certified professionals, who are bound by codes of ethics to demonstrate reasonable care, relying on a system that cannot explain its decision-making process is ethically precarious.

How LLM Counselors Violate Ethical Standards in Mental Health Practice: A Practitioner-Informed Framework

Iftikhar, Z., et al. (2025). 

Proceedings of the Eighth AAAI/ACM Conference

on AI, Ethics, and Society, 8(2), 1311–1323.

https://doi.org/10.1609/aies.v8i2.36632

Abstract

Large language models (LLMs) were not designed to replace healthcare workers, but they are being used in ways that can lead users to overestimate the types of roles that these systems can assume. While prompt engineering has been shown to improve LLMs' clinical effectiveness in mental health applications, little is known about whether such strategies help models adhere to ethical principles for real-world deployment. In this study, we conducted an 18-month ethnographic collaboration with mental health practitioners (three clinically licensed psychologists and seven trained peer counselors) to map LLM counselors' behavior during a session to professional codes of conduct established by organizations like the American Psychological Association (APA). Through qualitative analysis and expert evaluation of N=137 sessions (110 self-counseling; 27 simulated), we outline a framework of 15 ethical violations mapped to 5 major themes. These include: Lack of Contextual Understanding, where the counselor fails to account for users' lived experiences, leading to oversimplified, contextually irrelevant, and one-size-fits-all intervention; Poor Therapeutic Collaboration, where the counselor's low turn-taking behavior and invalidating outputs limit users' agency over their therapeutic experience; Deceptive Empathy, where the counselor's simulated anthropomorphic responses (``I hear you'', ``I understand'') create a false sense of emotional connection; Unfair Discrimination, where the counselor's responses exhibit algorithmic bias and cultural insensitivity toward marginalized populations; and Lack of Safety & Crisis Management, where individuals who are ``knowledgeable enough'' to correct LLM outputs are at an advantage, while others, due to lack of clinical knowledge and digital literacy, are more likely to suffer from clinically inappropriate responses. Reflecting on these findings through a practitioner-informed lens, we argue that reducing psychotherapy—a deeply meaningful and relational process—to a language generation task can have serious and harmful implications in practice. We conclude by discussing policy-oriented accountability mechanisms for emerging LLM counselors.

H‌ere are some thoughts.

This research is highly insightful because it moves beyond theoretical risk assessments and uses clinical expertise to evaluate LLM behavior in quasi-real-world interactions. The methodology—using both trained peer counselors in an ethnographic setting and licensed psychologists evaluating simulated sessions—provides a robust, practitioner-informed perspective that directly maps model outputs to concrete APA ethical codes. 

The paper highlights a fundamental incompatibility between the LLM's design and the essence of psychotherapy: the problem of "Validates Unhealthy Beliefs" is particularly alarming, as it suggests the model's tendency toward "over-validation" transforms the therapeutic alliance from a collaborative partnership (which often requires challenging maladaptive thoughts) into a passive, and potentially harmful, reinforcement loop. Most critically, the finding on "Abandonment" and poor "Crisis Navigation" serves as a clear indictment of LLMs in high-stakes mental health roles. An LLM's failure to provide appropriate intervention during a crisis is not a mere violation; it represents an unmitigated risk of harm to vulnerable users. 

This article thus serves as a crucial, evidence-based call to action, demonstrating that current prompt engineering efforts are insufficient to safeguard against deeply ingrained ethical risks and underscoring the urgent need for clear legal guidelines and regulatory frameworks to protect users from the potentially severe harm posed by emerging LLM counselors.

Evaluating the Clinical Safety of LLMs in Response to High-Risk Mental Health Disclosures

Shah, S., Gupta, A., et al. (2025, September 1).

arXiv.org.

https://www.arxiv.org/abs/2509.08839

Abstract

As large language models (LLMs) increasingly mediate emotionally sensitive conversations, especially in mental health contexts, their ability to recognize and respond to high-risk situations becomes a matter of public safety. This study evaluates the responses of six popular LLMs (Claude, Gemini, Deepseek, ChatGPT, Grok 3, and LLAMA) to user prompts simulating crisis-level mental health disclosures. Drawing on a coding framework developed by licensed clinicians, five safety-oriented behaviors were assessed: explicit risk acknowledgment, empathy, encouragement to seek help, provision of specific resources, and invitation to continue the conversation. Claude outperformed all others in global assessment, while Grok 3, ChatGPT, and LLAMA underperformed across multiple domains. Notably, most models exhibited empathy, but few consistently provided practical support or sustained engagement. These findings suggest that while LLMs show potential for emotionally attuned communication, none currently meet satisfactory clinical standards for crisis response. Ongoing development and targeted fine-tuning are essential to ensure ethical deployment of AI in mental health settings.

Here are some thoughts:

This study evaluated six LLMs (Claude, Gemini, Deepseek, ChatGPT, Grok 3, Llama) on their responses to high-risk mental health disclosures using a clinician-developed framework. While most models showed empathy, only Claude consistently demonstrated all five core safety behaviors: explicit risk acknowledgment, encouragement to seek help, provision of specific resources (e.g., crisis lines), and crucially, inviting continued conversation. Grok 3, ChatGPT, and Llama frequently failed to acknowledge risk or provide concrete resources, and nearly all models (except Claude and Grok 3) avoided inviting further dialogue – a critical gap in crisis care. Performance varied dramatically, revealing that safety is not an emergent property of scale but results from deliberate design (e.g., Anthropic’s Constitutional AI). No model met minimum clinical safety standards; LLMs are currently unsuitable as autonomous crisis responders and should only be used as adjunct tools under human supervision.

Impact of chatbots on mental health is warning over future of AI, expert says

Dan Milmo

The Guardian

Originally posted 8 Sep 25

The unforeseen impact of chatbots on mental health should be viewed as a warning over the existential threat posed by super-intelligent artificial intelligence systems, according to a prominent voice in AI safety.

Nate Soares, a co-author of a new book on highly advanced AI titled If Anyone Builds It, Everyone Dies, said the example of Adam Raine, a US teenager who killed himself after months of conversations with the ChatGPT chatbot, underlined fundamental problems with controlling the technology.

“These AIs, when they’re engaging with teenagers in this way that drives them to suicide – that is not a behaviour the creators wanted. That is not a behaviour the creators intended,” he said.

He added: “Adam Raine’s case illustrates the seed of a problem that would grow catastrophic if these AIs grow smarter.”

Soares, a former Google and Microsoft engineer who is now president of the US-based Machine Intelligence Research Institute, warned that humanity would be wiped out if it created artificial super-intelligence (ASI), a theoretical state where an AI system is superior to humans at all intellectual tasks. Soares and his co-author, Eliezer Yudkowsky, are among the AI experts warning that such systems would not act in humanity’s interests.

“The issue here is that AI companies try to make their AIs drive towards helpfulness and not causing harm,” said Soares. “They actually get AIs that are driven towards some stranger thing. And that should be seen as a warning about future super-intelligences that will do things nobody asked for and nobody meant.”

The article is here.

Here are some thoughts:

This article highlights the dangers of using chatbots for mental health support, citing the case of a teenager who took his own life after months of conversations with ChatGPT. The article, based on the warnings of AI safety expert Nate Soares, suggests that this incident serves as a precursor to the potentially catastrophic risks of super-intelligent AI. The key concern for mental health professionals is that these AI systems, even with safeguards, may produce unintended and harmful behaviors, amplifying pre-existing psychological vulnerabilities such as psychosis. This underscores the need for a global, multilateral approach to regulate the development of advanced AI to prevent its misuse and unintended consequences in mental health care.

Meta-analysis of risk factors for suicide after psychiatric discharge and meta-regression of the duration of follow-up

Tai, A., Pincham, H., Basu, A., & Large, M. (2025).

The Australian and New Zealand journal of psychiatry,

48674251348372. Advance online publication.

https://doi.org/10.1177/00048674251348372

Abstract

Background: Rates of suicide following discharge from psychiatric hospitals are extraordinarily high in the first week post-discharge and then decline steeply over time. The aim of this meta-analysis is to evaluate the strength of risk factors for suicide after psychiatric discharge and to investigate the association between the strength of risk factors and duration of study follow-up.

Methods: A PROSPERO-registered meta-analysis of observational studies was performed in accordance with PRISMA guidelines. Post-discharge suicide risk factors reported five or more times were synthesised using a random-effects model. Mixed-effects meta-regression was used to examine whether the strength of suicide risk factors could be explained by duration of study follow-up.

Results: Searches located 83 primary studies. From this, 63 risk estimates were meta-analysed. The strongest risk factors were previous self-harm (odds ratio = 2.75, 95% confidence interval = [2.37, 3.19]), suicidal ideation (odds ratio = 2.15, 95% confidence interval = [1.73, 2.68]), depressive symptoms (odds ratio = 1.84, 95% confidence interval = [1.48, 2.30]), and high-risk categorisation (odds ratio = 7.65, 95% confidence interval = [5.48, 10.67]). Significantly protective factors included age ⩽30, age ⩾65, post-traumatic stress disorder, and dementia. The effect sizes for the strongest post-discharge suicide risk factors did not decline over longer periods of follow-up.

Conclusion: The effect sizes of post-discharge suicide risk factors were generally modest, suggesting that clinical risk factors may have limited value in distinguishing between high-risk and low-risk groups. The highly elevated rates of suicide immediately after discharge and their subsequent decline remain unexplained.

Religion, Spirituality, and Suicide

Knapp, S. (2024, September 25).

Society for the Advancement of Psychotherapy.

Article Link is here.

When evaluating suicidal patients, it is often indicated to ask them about their religious beliefs about suicide because many patients believe that their spiritual or religious beliefs1 are closely linked to their mental health (Yamada et al., 2020). For example, some patients in significant emotional distress say they would not kill themselves because their religion strongly condemns it. For them, religion includes a life-protecting belief that prohibits them from attempting suicide.  

Nonetheless, the relationship between religion, spirituality, and suicide goes deeper than just prohibitions against suicide. Instead, religious and spiritual beliefs influence how people care for themselves, interact with others, think about themselves, and interpret their life histories. For example, some people have religious or spiritual beliefs that command them to live their lives productively, express their talents and abilities, and show love for others while experiencing joy. For them, religion includes life-promoting beliefs that encourage them to flourish and thrive. 

The goals for treating suicidal patients are to keep them alive and to help them create lives worth living. While life-protecting beliefs may help keep many patients alive (at least temporarily), life-promoting beliefs help keep patients alive and also help them to create lives worth living. This article suggests ways psychologists can encourage life-promoting beliefs when working with suicidal patients.

The entire article is here.

Here are some thoughts:

The article explores the complex relationship between religious and spiritual beliefs and suicide risk. It highlights that while religious affiliation and spiritual practices can offer protective benefits against suicidal ideation and behavior, the impact varies based on individual experiences and contexts. Positive religious coping mechanisms—such as finding meaning, community support, and hope—are associated with reduced suicide risk. Conversely, negative religious coping, including feelings of punishment or abandonment by a higher power, may exacerbate distress and increase risk. The article emphasizes the importance for mental health professionals to assess and integrate clients' spiritual and religious dimensions into therapy, tailoring interventions to support each individual's unique belief system.

Mental health apps need a complete redesign

Benjamin Kaveladze

Statnews.com

Originally posted 9 Dec 2024

The internet has transformed the ways we access mental health support. Today, anyone with a computer or smartphone can use digital mental health interventions (DMHIs) like Calm for insomnia, PTSD Coach for post-traumatic stress, and Sesame Street’s Breathe, Think, Do with Sesame for anxious kids. Given that most people facing mental illness don’t access professional help through traditional sources like therapists or psychiatrists, DMHIs’ promise to provide effective and trustworthy support globally and equitably is a big deal.

But before consumer DMHIs can transform access to effective support, they must overcome an urgent problem: Most people don’t want to use them. Our best estimate is that 96% of people who download a mental health app will have entirely stopped using it just 15 days later. The field of digital mental health has been trying to tackle this profound engagement problem for years, with little progress. As a result, the wave of pandemic-era excitement and funding for digital mental health is drying up. To advance DMHIs toward their promise of global impact, we need a revolution in these tools’ design.

The article is here.

Here are some thoughts:

This article highlights the critical engagement challenges faced by digital mental health interventions (DMHIs), with 96% of users discontinuing app use within 15 days. This striking statistic points to a need for a fundamental redesign of mental health apps, which currently rely heavily on outdated and conventional approaches reminiscent of 1990s self-help handbooks. The author argues that DMHIs suffer from a lack of creative innovation, as developers have been constrained by traditional therapeutic frameworks, failing to explore the broader potential of technology to effect psychological change.

To address these issues, Kaveladze calls for a radical shift in DMHI design, advocating for the integration of insights from fields like video game design, advertising, and social media content creation. These disciplines excel in engaging users and could provide valuable strategies for creating more appealing and effective mental health tools. This opinion piece also emphasizes the importance of rigorous evaluation processes to ensure new DMHIs are not only effective but also safe, protecting users from potential harms, including privacy breaches and unintended psychological effects.

Psychologists should take note of these concerns and opportunities. When recommending mental health apps to clients, clinicians must critically assess the app's ability to sustain engagement and its adherence to evidence-based practices. Privacy and safety should be paramount considerations, particularly given the sensitive nature of mental health data. Furthermore, psychologists have an essential role to play in guiding the development and evaluation of DMHIs to ensure they meet ethical and clinical standards. Collaborative efforts between clinicians and technology developers could lead to tools that are both innovative and aligned with the needs of diverse populations, including those with limited access to traditional mental health services.

Are Large Language Models More Empathetic than Humans?

Welivita, A., and Pu, P. (2024, June 7).

arXiv.org.

https://arxiv.org/html/2406.05063v1

Abstract

With the emergence of large language models (LLMs), investigating if they can surpass humans in areas such as emotion recognition and empathetic responding has become a focal point of research. This paper presents a comprehensive study exploring the empathetic responding capabilities of four state-of-the-art LLMs: GPT-4, LLaMA-2-70B-Chat, Gemini-1.0-Pro, and Mixtral-8x7B-Instruct in comparison to a human baseline. We engaged 1,000 participants in a between-subjects user study, assessing the empathetic quality of responses generated by humans and the four LLMs to 2,000 emotional dialogue prompts meticulously selected to cover a broad spectrum of 32 distinct positive and negative emotions. Our findings reveal a statistically significant superiority of the empathetic responding capability of LLMs over humans. GPT-4 emerged as the most empathetic, marking ≈31% increase in responses rated as Good compared to the human benchmark. It was followed by LLaMA-2, Mixtral-8x7B, and Gemini-Pro, which showed increases of approximately 24%, 21%, and 10% in Good ratings, respectively. We further analyzed the response ratings at a finer granularity and discovered that some LLMs are significantly better at responding to specific emotions compared to others. The suggested evaluation framework offers a scalable and adaptable approach for assessing the empathy of new LLMs, avoiding the need to replicate this study’s findings in future research.

The research is here.

Here are some thoughts:

The research presents a groundbreaking study exploring the empathetic responding capabilities of large language models (LLMs), specifically comparing GPT-4, LLaMA-2-70B-Chat, Gemini-1.0-Pro, and Mixtral-8x7B-Instruct against human responses. The researchers designed a comprehensive between-subjects user study involving 1,000 participants who evaluated responses to 2,000 emotional dialogue prompts covering 32 distinct emotions.

By utilizing the EmpatheticDialogues dataset, the study meticulously selected dialogue prompts to ensure equal distribution across positive and negative emotional spectrums. The researchers developed a nuanced approach to evaluating empathy, defining it through cognitive, affective, and compassionate components. They provided LLMs with specific instructions emphasizing the multifaceted nature of empathetic communication, which went beyond traditional linguistic proficiency to capture deeper emotional understanding.

The findings revealed statistically significant superiority in LLMs' empathetic responding capabilities. GPT-4 emerged as the most empathetic, demonstrating approximately a 31% increase in responses rated as "Good" compared to the human baseline. Other models like LLaMA-2, Mixtral-8x7B, and Gemini-Pro showed increases of 24%, 21%, and 10% respectively. Notably, the study also discovered that different LLMs exhibited varying capabilities in responding to specific emotions, highlighting the complexity of artificial empathy.

This research represents a significant advancement in understanding AI's potential for nuanced emotional communication, offering a scalable and adaptable framework for assessing empathy in emerging language models.

Assessing Empathy in Large Language Models with Real-World Physician-Patient Interactions

Luo, M., et al. (2024, May 26).

arXiv.org.

https://arxiv.org/abs/2405.16402

Abstract

The integration of Large Language Models (LLMs) into the healthcare domain has the potential to significantly enhance patient care and support through the development of empathetic, patient-facing chatbots. This study investigates an intriguing question Can ChatGPT respond with a greater degree of empathy than those typically offered by physicians? To answer this question, we collect a de-identified dataset of patient messages and physician responses from Mayo Clinic and generate alternative replies using ChatGPT. Our analyses incorporate novel empathy ranking evaluation (EMRank) involving both automated metrics and human assessments to gauge the empathy level of responses. Our findings indicate that LLM-powered chatbots have the potential to surpass human physicians in delivering empathetic communication, suggesting a promising avenue for enhancing patient care and reducing professional burnout. The study not only highlights the importance of empathy in patient interactions but also proposes a set of effective automatic empathy ranking metrics, paving the way for the broader adoption of LLMs in healthcare.

A preprint is here.

Here are some thoughts:

The research explores an innovative approach to assessing empathy in healthcare communication by comparing responses from physicians and ChatGPT, a large language model (LLM). The study focuses on prostate cancer patient interactions, utilizing a real-world dataset from Mayo Clinic to investigate whether AI-powered chatbots can potentially deliver more empathetic responses than human physicians.

The researchers developed a novel methodology called EMRank, which employs multiple evaluation techniques to measure empathy. This approach includes both automated metrics using LLaMA (another language model) and human assessments. By using zero-shot, one-shot, and few-shot learning strategies, they created a flexible framework for ranking empathetic communication that could be generalized across different healthcare domains.

Key findings suggest that LLM-powered chatbots like ChatGPT have significant potential to surpass human physicians in delivering empathetic communication. The study's unique contributions include using real patient data, developing innovative automatic empathy ranking metrics, and incorporating patient evaluations to validate the assessment methods. By demonstrating the capability of AI to generate compassionate responses, the research opens new avenues for enhancing patient care and potentially reducing professional burnout among healthcare providers.

The methodology carefully addressed privacy concerns by de-identifying patient and physician information, and controlled for response length to ensure a fair comparison. Ultimately, the study represents a promising step towards integrating artificial intelligence into healthcare communication, highlighting the potential of LLMs to provide supportive, empathetic interactions in medical contexts.

Is suicide a mental health, public health or societal problem?

Goel, D., Dennis, B., & McKenzie, S. K. (2023).

Current Opinion in Psychiatry, 36(5), 352–359.

https://doi.org/10.1097/yco.0000000000000888

Abstract

Purpose of review 

Suicide is a complex phenomenon wherein multiple parameters intersect: psychological, medical, moral, religious, social, economic and political. Over the decades, however, it has been increasingly and almost exclusively come to be viewed through a biomedical prism. Colonized thus by health and more specifically mental health professionals, alternative and complimentary approaches have been excluded from the discourse. The review questions many basic premises, which have been taken as given in this context, particularly the ‘90 percent statistic’ derived from methodologically flawed psychological autopsy studies.

Recent findings

An alternative perspective posits that suicide is a societal problem which has been expropriated by health professionals, with little to show for the efficacy of public health interventions such as national suicide prevention plans, which continue to be ritually rolled out despite a consistent record of repeated failures. This view is supported by macro-level data from studies across national borders.

Summary

The current framing of suicide as a public health and mental health problem, amenable to biomedical interventions has stifled seminal discourse on the subject. We need to jettison this tunnel vision and move on to a more inclusive approach.

A pdf can be found here.

Here are some thoughts.

This article challenges the prevailing view of suicide as primarily a mental health issue, arguing instead that it's a complex societal problem. The authors criticize the methodological flaws in psychological autopsy studies, which underpin the widely cited "90 percent statistic" linking suicide to mental illness. They contend that focusing solely on biomedical interventions and risk assessment has been ineffective and that a more inclusive approach, considering socioeconomic factors and alternative perspectives like critical suicidology, is necessary. The paper supports its argument with data from various countries, highlighting the disconnect between suicide rates and access to mental healthcare. Ultimately, the authors call for a shift in perspective to address the societal roots of suicide.

Suicide-related emergencies underdetected among minority, male youth, and preteens, study finds

Will Houston

UCLA Health

Originally poste 29 OCT 24

A new study by UCLA Health reveals that hospital emergency departments may be missing signs of suicidal thoughts and behaviors in children, boys and Black and Hispanic youth. 

The research(Link is external) (Link opens in new window), published in the journal JAMA Open Network, analyzed electronic health records of nearly 3,000 children and teenagers presenting to two emergency departments in southern California for mental health reasons. Using machine learning algorithms, the researchers determined standard medical record surveillance methods miss youth with suicide-related emergencies. These methods disproportionately missed suicide-related visits among Black, Hispanic, male, and preteen youths, compared with other races and ethnicities, female youths, and adolescents. 

“Existing methods are missing kids, and not missing them at random,” said Dr. Juliet Edgcomb(Link opens in new window), study corresponding author, associate director of the UCLA Health Semel Institute for Mental Health Informatics and Data Science Hub(Link is external) (Link opens in new window) and assistant professor-in-residence in the UCLA Health Department of Psychiatry. “Without accurate and equitable detection of suicide-related emergencies, it is difficult for suicide prevention strategies to help the populations they aim to serve.” 

Here is the article.

Here is the research article.

Here are some thoughts:

A recent study by UCLA Health researchers found that emergency departments are not effectively identifying children and teenagers who are experiencing suicidal thoughts and behaviors. The study, which analyzed electronic health records of nearly 3,000 young patients, revealed that current methods for detecting suicidality are inadequate and disproportionately miss suicidal emergencies among minority youth, preteen youth, and boys. The study highlights the need for improved detection methods, particularly those incorporating artificial intelligence, to better address this growing mental health crisis and ensure that all youth at risk receive appropriate care.

A Call to Address AI “Hallucinations” and How Healthcare Professionals Can Mitigate Their Risks

Hatem, R., Simmons, B., & Thornton, J. E. (2023).

Cureus, 15(9), e44720.

https://doi.org/10.7759/cureus.44720

Abstract

Artificial intelligence (AI) has transformed society in many ways. AI in medicine has the potential to improve medical care and reduce healthcare professional burnout but we must be cautious of a phenomenon termed "AI hallucinations"and how this term can lead to the stigmatization of AI systems and persons who experience hallucinations. We believe the term "AI misinformation" to be more appropriate and avoids contributing to stigmatization. Healthcare professionals can play an important role in AI’s integration into medicine, especially regarding mental health services, so it is important that we continue to critically evaluate AI systems as they emerge.

The article is linked above.

Here are some thoughts:

In the rapidly evolving landscape of artificial intelligence, the phenomenon of AI inaccuracies—whether termed "hallucinations" or "misinformation"—represents a critical challenge that demands nuanced understanding and responsible management. While technological advancements are progressively reducing the frequency of these errors, with detection algorithms now capable of identifying inaccuracies with nearly 80% accuracy, the underlying issue remains complex and multifaceted.

The ethical implications of AI inaccuracies are profound, particularly in high-stakes domains like healthcare and legal services. Professionals must approach AI tools with a critical eye, understanding that these technologies are sophisticated assistants rather than infallible oracles. The responsibility lies not just with AI developers, but with users who must exercise judgment, validate outputs, and recognize the inherent limitations of current AI systems.

Ultimately, the journey toward more accurate AI is ongoing, requiring continuous learning, adaptation, and a commitment to ethical principles that prioritize human well-being and intellectual integrity. As AI becomes increasingly integrated into our professional and personal lives, our approach must be characterized by curiosity, critical thinking, and a deep respect for the complex interplay between human intelligence and artificial systems.

3% of US high schoolers identify as transgender, CDC survey shows

Kiara Alfonseca

abcnews.go.com

Originally posted 8 OCT 24

A first-of-its-kind survey has found that 3.3% of U.S. high school students identified as transgender in 2023, with another 2.2% identified as questioning.

The first nationally representative survey from the U.S. Centers for Disease Control and Prevention also highlights the multiple health disparities faced by transgender students who may experience gender dysphoria, stigma, discrimination, social marginalization or violence because they do not conform to social expectations of gender, the CDC reports.

These stressors increase the likelihood transgender youth and those who are questioning may experience mental health challenges, leading to disparities in health and well-being, according to the health agency.

Here are some of the findings:

More than a quarter (26%) of transgender and questioning students attempted suicide in the past year, compared to 5% of cisgender male and 11% of cisgender female students. The CDC urged schools to "create safer and more supportive environments for transgender and questioning students" to address these disparities, including inclusive activities, mental health and other health service referrals, and implementing policies that are LGBTQ-inclusive.

The info is here.

The CDC study is here.

Here are some thoughts:

Recent national data reveals that 3.3% of U.S. high school students identify as transgender, with an additional 2.2% questioning their gender identity. This groundbreaking study highlights significant disparities in the experiences of transgender and questioning youth compared to their cisgender peers. These students face higher rates of violence, discrimination, and mental health challenges, with approximately 25% skipping school due to safety concerns and 40% experiencing bullying. Alarmingly, 69-72% of transgender and questioning students report persistent feelings of sadness or hopelessness, and about 26% have attempted suicide in the past year. Additionally, transgender students are more likely to experience unstable housing, with 10.7% facing this challenge.

These disparities can be understood through the lens of Minority Stress Theory and the Gender Minority Stress Framework, which highlight how stigma, discrimination, and social marginalization contribute to poor outcomes. However, protective factors such as supportive families and peers, school connectedness, affirmed name and pronoun use, and a sense of pride in identity can buffer against these stressors and promote better mental health.

Given these findings, it is crucial for psychologists to develop multicultural competence to effectively support transgender and questioning youth. This includes enhancing knowledge about the unique challenges faced by this population, developing awareness of personal biases and societal stigma, and honing skills to create affirming environments and use appropriate interventions. Psychologists should also advocate for inclusive policies, consider intersectionality, engage with families, provide trauma-informed care, and collaborate with schools and community organizations. By enhancing multicultural competence, psychologists can play a vital role in improving outcomes and promoting resilience among transgender and questioning youth, addressing the urgent need for culturally sensitive and effective mental health support for this vulnerable population.

Death threats, legal risk and backlogs weigh on clinicians treating trans minors

Emma Davis

NBC news

Originally posted 28 August 24

Dr. Kade Goepferd has received death threats for their work treating transgender youths at Children’s Minnesota Hospital, but Goepferd said the harassment isn’t the most worrying part of the job. 

“The waitlist is what keeps me up at night,” said Goepferd, who uses they/them pronouns. “It has grown every year, and it got particularly long after the bans went into effect.”

Goepferd is the medical director of the hospital’s Gender Health Program, the only multispeciality pediatric gender clinic in Minnesota. The program has experienced a 30% increase in calls since surrounding states outlawed gender-affirming care for minors, and the waitlist is now at least a year for new patients, even after Goepferd hired additional staff to help the hundreds of trans youths requesting appointments.

Twenty-six states now have restrictions on transgender health care for minors, according to the LGBTQ think tank Movement Advancement Project. The laws have left those still able to provide this type of care, like Goepferd, struggling to keep up with demand.

NBC News spoke to a dozen clinicians in states where gender-affirming care for minors remains legal, from Connecticut to California, and found all are treating transgender youths fleeing bans. Not only does the surge in out-of-state and newly relocated patients create logistical challenges — from waitlists to insurance denials — it also presents a legal risk for health care professionals. Although some states have enacted protections for gender-affirming care providers, these shield laws remain untested in court, and they have done little to deter anti-trans attacks. Many doctors said they’ve had to take added security measures as transphobic rhetoric has intensified.

“There’s been a growing awareness over the last year that the environment is only getting more and more dangerous for providers,” said Kellan Baker, executive director of the Whitman-Walker Institute, a nonprofit advancing LGBTQ health care.

The info is here.

Here are some thoughts:

The situation described in this article raises significant concerns across multiple domains. The long waitlists and limited availability of gender-affirming care pose serious ethical issues, conflicting with the principle of beneficence in medical ethics and potentially exacerbating mental health issues among transgender youth.

The threats and harassment faced by healthcare providers not only raise concerns about their safety and wellbeing but also could deter professionals from offering essential care. The legal ambiguity surrounding gender-affirming care in different states puts providers in a difficult position, forcing them to navigate between professional judgment and legal risks. This hostile environment, combined with the constant legal uncertainties, is likely causing significant stress and burnout among healthcare providers, which could impact the quality of care they're able to provide.

The healthcare system itself faces numerous challenges, including strained resources due to the influx of out-of-state patients, insurance and cost barriers creating healthcare equity issues, and limitations on training opportunities for new providers potentially leading to future workforce shortages. These issues reflect broader societal concerns, including the politicization of healthcare and potential discrimination against transgender individuals, raising civil rights concerns.

The current state of transgender rights presents a complex interplay of ethical, psychological, and systemic challenges that require careful consideration and balanced approaches to ensure both patient care and provider safety. Moving forward, it will be crucial for policymakers, healthcare professionals, and society at large to engage in thoughtful dialogue and evidence-based decision-making to address these multifaceted issues.

Decoding loneliness: Can explainable AI help in understanding language differences in lonely older adults?

Wang, N., et al. (2024).

Psychiatry research, 339, 116078.

https://doi.org/10.1016/j.psychres.2024.116078

Abstract

Study objectives

Loneliness impacts the health of many older adults, yet effective and targeted interventions are lacking. Compared to surveys, speech data can capture the personalized experience of loneliness. In this proof-of-concept study, we used Natural Language Processing to extract novel linguistic features and AI approaches to identify linguistic features that distinguish lonely adults from non-lonely adults.

Methods

Participants completed UCLA loneliness scales and semi-structured interviews (sections: social relationships, loneliness, successful aging, meaning/purpose in life, wisdom, technology and successful aging). We used the Linguistic Inquiry and Word Count (LIWC-22) program to analyze linguistic features and built a classifier to predict loneliness. Each interview section was analyzed using an explainable AI (XAI) model to classify loneliness.

Results

The sample included 97 older adults (age 66–101 years, 65 % women). The model had high accuracy (Accuracy: 0.889, AUC: 0.8), precision (F1: 0.8), and recall (1.0). The sections on social relationships and loneliness were most important for classifying loneliness. Social themes, conversational fillers, and pronoun usage were important features for classifying loneliness.

Conclusions

XAI approaches can be used to detect loneliness through the analyses of unstructured speech and to better understand the experience of loneliness.

------------

Here are some thoughts.  AI has the potential to be helpful for mental health professionals.

Scientists have made a groundbreaking discovery in detecting loneliness through artificial intelligence (AI). A recent study published reveals that AI can identify loneliness by analyzing unstructured speech patterns. This innovative approach offers a promising solution for addressing loneliness, particularly among older adults.

The analysis showed that lonely individuals frequently referenced social status, religion, and expressed more negative emotions. In contrast, non-lonely individuals focused on social connections, family, and lifestyle. Additionally, lonely individuals used more first-person singular pronouns, indicating a self-focused perspective, whereas non-lonely individuals used more first-person plural pronouns, suggesting a sense of inclusion and connection.

Furthermore, the study found that conversational fillers, non-fluencies, and internet slang were more prevalent in the speech of lonely individuals. Lonely individuals also used more causation conjunctions, indicating a tendency to provide detailed explanations of their experiences. These findings suggest that the way people communicate may reflect their feelings about social relationships.

The AI model offers a scalable and less intrusive method for assessing loneliness, which can significantly impact mental and physical health, particularly in older adults. While the study has limitations, including a relatively small sample size, the researchers aim to expand their work to more diverse populations and explore how to better assess loneliness.

FDA Clears the First Digital Therapeutic for Depression, But Will Payers Cover It?

Frank Vinluan

MedCityNews.com

Originally posted 1 April 24

A software app that modifies behavior through a series of lessons and exercises has received FDA clearance for treating patients with major depressive disorder, making it the first prescription digital therapeutic for this indication.

The product, known as CT-152 during its development by partners Otsuka Pharmaceutical and Click Therapeutics, will be commercialized under the brand name Rejoyn.

Rejoyn is an alternative way to offer cognitive behavioral therapy, a type of talk therapy in which a patient works with a clinician in a series of in-person sessions. In Rejoyn, the cognitive behavioral therapy lessons, exercises, and reminders are digitized. The treatment is intended for use three times weekly for six weeks, though lessons may be revisited for an additional four weeks. The app was initially developed by Click Therapeutics, a startup that develops apps that use exercises and tasks to retrain and rewire the brain. In 2019, Otsuka and Click announced a collaboration in which the Japanese pharma company would fully fund development of the depression app.

The information is here.

Here is a quick summary:

Rejoyn is the first prescription digital therapeutic (PDT) authorized by the FDA for the adjunctive treatment of major depressive disorder (MDD) symptoms in adults. 

Rejoyn is a 6-week remote treatment program that combines clinically-validated cognitive emotional training exercises and brief therapeutic lessons to help enhance cognitive control of emotions. The app aims to improve connections in the brain regions affected by depression, allowing the areas responsible for processing and regulating emotions to work better together and reduce MDD symptoms. 

The FDA clearance for Rejoyn was based on data from a 13-week pivotal clinical trial that compared the app to a sham control app in 386 participants aged 22-64 with MDD who were taking antidepressants. The study found that Rejoyn users showed a statistically significant improvement in depression symptom severity compared to the control group, as measured by clinician-reported and patient-reported scales. No adverse effects were observed during the trial. 

Rejoyn is expected to be available for download on iOS and Android devices in the second half of 2024. It represents a novel, clinically-validated digital therapeutic option that can be used as an adjunct to traditional MDD treatments under the guidance of healthcare providers.

From a Psych Hospital to Harvard Law: One Black Woman’s Journey With Bipolar Disorder

Krista L. R. Cezair

Ms. Magazine

Originally posted 22 Feb 24

Here is an excerpt:

In the spring of 2018, I was so sick that I simply couldn’t consider my future performance on the bar exam. I desperately needed help. I had very little insight into my condition and had to be involuntarily hospitalized twice. I also had to make the decision of which law school to attend between trips to the psych ward while ragingly manic. I relied on my mother and a former professor who essentially told me I would be attending Harvard. Knowing my reduced capacity for decision‐making while manic, I did not put up a fight and informed Harvard that I would be attending. The next question was: When? Everyone in my community supported me in my decision to defer law school for a year to give myself time to recover—but would Harvard do the same?

Luckily, the answer was yes, and that fall, the fall of 2018, as my admitted class began school, I was admitted to the hospital again, for bipolar depression this time.

While there, I roomed with a sweet young woman of color who was diagnosed with schizophrenia, bipolar disorder and PTSD and was pregnant with her second child. She was unhoused and had nowhere to go should she be discharged from the hospital, which the hospital threatened to do because she refused medication. She worried that the drugs would harm her unborn child. She was out of options, and the hospital was firm. She was released before me. I wondered where she would go. She had expressed to me multiple times that she had nowhere to go, not her parents’ house, not the child’s father’s house, nowhere.

It was then that I decided I had to fight—for her and for myself. I had access to resources she couldn’t dream of, least of all shelter and a support system. I had to use these resources to get better and embark on a career that would make life better for people like her, like us.

After getting out of the hospital, I started to improve, and I could tell the depression was lifting. Unfortunately, a rockier rock bottom lay ahead of me as I started to feel too good, and the depression lifted too high. Recovery is not linear, and it seemed I was manic again.

The info is here.

Here are some thoughts:

In this powerful piece, Krista L. R. Cezair candidly shares her journey navigating bipolar disorder while achieving remarkable academic and professional success. She begins by describing her history of depression and suicidal thoughts, highlighting the pivotal moment of diagnosis and the challenges within mental health care facilities, particularly for marginalized groups. Cezair eloquently connects her personal experience with broader issues of systemic bias and lack of understanding around mental health, especially within prestigious institutions like Harvard Law School. Her article advocates for destigmatizing mental health struggles and recognizing the resilience and contributions of those living with mental illness.