Digital Mental Health for Older Adults: Foe or Friend?


The COVID-19 pandemic accelerated the integration of telemental health services for older adults. But such digitization also created new complexities regarding equitable access, privacy, and regulation. This article summarizes the evolving digital mental health landscape, focusing on regulatory frameworks, inherent risks, and the transformative potential of technology, particularly artificial intelligence (AI). Also covered are shifts in Medicare coverage and calls for standardized guidelines, interdisciplinary collaboration and robust regulation focused on patient privacy and AI ethics. Despite challenges, AI-driven diagnostic, monitoring, and intervention tools offer insights into mental health trajectories, using data generated by wearables, social media, and immersive technologies.

Key Words:

telemental health, digitization, regulatory frameworks, AI, wearables, social media data, immersive technology


The COVID-19 pandemic forced a rapid transition to telemental health services for older adults to meet rising demand. This was especially critical for rural providers due to distancing measures compounding existing access issues. While telemental health allowed for continuity of care, there were challenges. Older adults faced a learning curve to access these services, with many initially struggling with the technology needed for telehealth and requiring help from family or caregivers. Concerns existed about older adults with cognitive impairments or limited digital literacy, but proper instruction and support enabled many to use telehealth successfully. The rapid shift also led to concerns about ethical, legal, and clinical repercussions for older adults and the importance of addressing these barriers (Kuang, et al., 2021; Mazziotti & Rutigliano, 2021).

The adoption of telemedicine also has spurred the integration of other technologies into the daily lives of older adults, partly aided by the increasing ubiquity of artificial intelligence (AI)-based tools. In this article, we dissect the risks and benefits associated with the adoption of technology for older adults. Technology may spark new ways in which we conceptualize and treat mental illness and open the way to a more inclusive, accessible, and ultimately beneficial mental healthcare for older adults. This must be balanced against real risks of further fragmentation in care, more exclusive access for some and lesser access for others, and the potential compromise of fundamental privacy and data security (Vahia et al., 2022).

In this article, written nearly 4 years after the COVID-19 pandemic, we consider the ways in which the digital mental health space for older adults is evolving. We first review the regulatory landscape for digital mental health and then consider the risks inherent with the use of various technologies. With these two aspects in context, we then review the potential of various technologies (primarily those that are based on AI) for diagnosis, monitoring, and intervention for older adults.

The Regulatory Landscape

Historically Medicare, state Medicaid programs, and private insurers have imposed limitations on coverage of telemedicine services (Jonk et al., 2020). The onset of the COVID-19 pandemic prompted a temporary relaxation of rules, allowing mental health professionals to provide services across state lines without additional licensure, and flexibility in HIPAA guidelines for telehealth privacy and security. Given the benefits documented during the pandemic, calls for standardized guidelines or laws regulating confidential information collection and AI in healthcare have grown, emphasizing patient privacy, safety, and ethical data use (Aggarwal et al., 2021; Jonk et al., 2020).

‘AI may be involved in decisions related to end-of-life care or advanced directives for older adults, leading to complex ethical dilemmas.’

Following the end of the healthcare emergency in May 2023, some flexibility offered by telemedicine was lost, but prior to this, the December 2022 Omnibus bill extended key supports, including Medicaid and Medicare coverage for telehealth services. These changes now grant access to behavioral/mental telehealth services, allowing patients to receive care at home without geographic restrictions, including the use of audio-only communication platforms. There is less regulation around other technologies, including AI in healthcare. The Food and Drug Administration, American Medical Association (AMA), and the World Health Organization (WHO) have proposed recommendations for ethical AI use, encompassing transparency, data privacy, safety, reliability, and accountability (Crigger et al., 2022). They also have proposed frameworks for regulation, including monitoring and validation of training data, transparency in algorithm development, and collaboration between stakeholders (WHO, ‎2023). However, these frameworks thus far are broad and do not specifically address mental health.

Inherent Risks:

As we review the integration of technology and AI in the mental healthcare of older adults, it is imperative to appreciate the inherent ethical, legal, and clinical risks. Below, we tabulate key considerations and potential challenges (Lieneck et al., 2020; Schriger et al., 2022; Tabibzadeh & Tran, 2022) that may emerge in that context.


Ethical Risks

Privacy and confidentiality

Concerns about the security of telehealth platforms and the risk of data breaches exposing sensitive patient information.

Informed consent

Patients may not fully understand the risks of telehealth, requiring clinicians to obtain thorough informed consent.

Standard of care

Concerns about delivering the same standard of care as in-person visits, especially for high-risk older adults or those with severe mental illness. 

Access issues

Not all older adults have access to the necessary technology or the digital literacy to use AI-driven mental health tools, thus exacerbating disparities in access to care.


AI may inherit biases in the datasets it uses, and inadvertently discriminate against older adults, as algorithms may not account for the unique challenges in aging.

End-of-life care

AI may be involved in decisions related to end-of-life care or advanced directives for older adults, leading to complex ethical dilemmas.

Legal Risks

Liability and malpractice

If an AI system provides harmful recommendations, it may lead to liability, potentially involving healthcare providers, developers, or healthcare institutions.

Reimbursement and funding

Legal and regulatory issues surrounding reimbursement and funding for telehealth services can pose challenges.

Licensing and jurisdiction

Licensing requirements and jurisdictional issues across state lines or international borders.

Patents and copyright

Creators of AI algorithms and software may need to address issues related to patents and copyright, especially if their AI systems have proprietary elements.

Documentation and record-keeping

Proper documentation and record-keeping are essential in telehealth to ensure compliance with legal and regulatory requirements.

Antitrust concerns

If a single AI provider or technology becomes dominant in AI-powered mental health for older adults, antitrust concerns may arise, potentially leading to legal action or regulatory intervention.

Data overload

AI can collect large amounts of patient data, ultimately leading to enhanced monitoring and personalization of healthcare. However, this may create a dilemma for clinicians with the potential risks if actionable data are missed.

Clinical Risks

Therapeutic relationship

Potential impact on the content and process of therapy as a result of altered engagement, retention, attendance, and varying digital literacy.

Crisis management

Challenges in evaluating and intervening with high-risk patients, including persons with suicidal thoughts.

Algorithmic errors

AI systems can make mistakes in diagnosing mental health conditions or providing treatment recommendations.

Lack of human expertise

AI may not fully replace the expertise and clinical judgment of human mental health professionals.


AI tools may focus on specific conditions or aspects, while neglecting to consider a person’s overall well-being and circumstances.


Reliance on AI may create dependance by patients and clinicians.


A New Suite of Clinical Tools:

The risks discussed above must be weighed against the range of tools that AI makes possible. These span diagnostics, monitoring, and intervention.

Diagnostic Tools:

An 81-year-old man with a persistent complaint of “brain fog” is evaluated by a psychiatrist after a medical and cognitive workup reveals no particular insight. The psychiatrist reviews the man’s medications and decides to taper off medications that could be sedating and contributing to this complaint. In addition to data provided by the patient, the psychiatrist also asks the patient to use a wearable device to track changes in sleep and activity during this process. The wearable data reveal that the man spends most of his time in bed. These data help the psychiatrist change the treatment plan to include non-pharmacologic behavioral activation approaches and switching to an activating antidepressant, in addition to eliminating potentially sedating medications. Over the next 6 weeks, the patient improves notably.

Increasingly, AI-based tools leverage diverse data sources, including electronic health records and behavioral data, to identify patterns and risk factors associated with mental health conditions in older adults. AI can discern subtle changes in behavior and symptoms and may be capable of predicting the onset of conditions like depression or dementia (Bartels et al., 2018). Moreover, AI can serve as a decision-support tool for healthcare providers, tailoring treatment plans based on individual needs by analyzing unique medical histories, genetics, and preferences (Ye & Liu, 2022). It also facilitates analysis of large existing datasets to identify trends that inform treatment strategies and healthcare policies.

‘Typing speed on smartphones has been shown to help in quantifying mood and cognition.’

Numerous examples exist in research of how these data can aid diagnostics. Typing speed on smartphones has been shown to help in quantifying mood and cognition (visual attention, processing speed, and task-switching) in people’s natural environment to complement formal assessments. A rapidly growing body of literature is demonstrating how sensors, either in standalone devices (e.g., the Emerald device developed at MIT) or complex sensor arrays (e.g., the platform developed at the Oregon Center for Aging & Technology) can capture a broad range of biomarkers of behavior and aging, laying the groundwork for personalized medicine (Au-Yeung et al., 2022; Zhang et al., 2021). The field of neuropsychology is leveraging data from web-based, open-access testing platforms (e.g., to create norms that are more inclusive than the current standard of cognitive testing (Singh & Germine, 2021).

An even more easily implementable tool is Ecological Momentary Assessment (EMA), which uses daily surveys and questionnaires to collect real-time data on individuals’ thoughts, feelings, behaviors, and experiences in their natural environments. Using EMA can help individuals and their clinicians track the day-to-day variations in mental health and function. It has helped shed light on multiple aspects of mental health, including triggers for changes in behavior, patterns of stress response and response to interventions, and has applications in diagnostics and monitoring (Dao et al., 2021).

Technology for Monitoring Symptoms

The integration of technology into mental health monitoring for older adults spans various devices and applications. Wearable devices, such as smartwatches and fitness trackers, gather data on physical activity, heart rate, and sleep patterns, with AI algorithms analyzing the information for signs of changes in mental well-being, such as decreased activity or sleep disruptions. Voice assistants and speech recognition systems contribute by analyzing speech patterns to detect emotional states, revealing conditions like anxiety, depression, or cognitive decline. Analysis of social media posts offers insights into users’ mental well-being based on language changes, post frequency, and shared content. Environmental sensors, including those in smart homes monitor factors like temperature, lighting, and movement patterns, correlating them with mental health changes (Chiu et al., 2020; Milnes-Ives et al., 2022; Pywell et al., 2020; Vijayan et al., 2021). Continuous monitoring at this degree of granularity can lead to important insights that facilitate personalized medicine.

There are, however, legitimate concerns among stakeholders about these approaches’ potential privacy infringement. A distinction must be made between monitoring and surveillance. While both involve continuous tracking of behaviors using technology, monitoring in the clinical domain implies tracking in the service of clinical decision making. Monitoring must rely on safeguards including informed consent for using such technologies, giving control to those being monitored over when to turn off the sensors to protect privacy and autonomy, and ensuring that data are collected only for periods when they may directly guide care decisions. Also, it is imperative that data be stored securely and deleted upon request of the patient or after they have served the clinical indication for which they were intended.

Technology in Intervention

A 69-year-old woman seeks support from a therapist to help cope with the sudden loss of her best friend. The therapist recommends that the woman also use an app in between sessions to practice mindfulness, as an additional resource to augment her psychotherapy sessions. The woman is initially unconvinced, but after a couple of attempts, notices that her heart rate seems to slow down from over 90 beats per minute to about 75 beats per minute after she spends a few minutes practicing the mindfulness exercise in the app. This motivates her to use the app more regularly. After 12 sessions, she completes her therapy successfully, but continues to use the app on occasion when she feels stressed.

‘Immersive Virtual Reality is finding a range of applications, including nature-based mindfulness-compassion programs and exposure therapies.’

Various technologies have found application in intervention delivery for older adults, particularly since the pandemic and scaling of telemedicine. Mobile applications coupled with wearables increasingly combine mood tracking, meditation, stress reduction, and personalized self-help tools catering to the unique needs and preferences of older adults. Supportive technologies enhance social support and facilitate participation in recreational activities (Similä et al., 2018). Online forums and communities reduce feelings of isolation by connecting older adults with peers. More recently, chatbots and virtual assistants have been shown to provide companionship, information, and therapeutic conversations. Robots like Paro offer emotional support and companionship, especially for those dealing with dementia and depression.

Immersive Virtual Reality (VR) is finding a range of applications, including nature-based mindfulness-compassion programs and exposure therapies (Skurla et al., 2022). Smartphones and tablet computers have been explored as cognitive and memory aids for older adults with and without cognitive impairment, and to promote general mental health. Reminiscence and Cognitive Therapy apps stimulate memory and cognitive function, proving beneficial for conditions like Alzheimer’s disease. AI-based cognitive training programs, tailored to the specific needs of older adults, hold potential for improving cognitive function and overall mental well-being (Wilson et al., 2022).

A novel self-administered intervention called “Let’s Talk Tech” for individuals with mild dementia and their caregivers, focuses on educating and facilitating communication about technologies used in dementia care. This intervention represents a new frontier by creating a loop where intervention delivery and quantification of cognitive/emotional function occur in lockstep (Berridge et al., 2023).

Projecting the Future:

The capacity for technologies to have a systemic impact on late life mental health will depend upon several factors. It will require substantial change in reimbursement policies to support the integration of technology. Outcome-based reimbursement models might present a viable solution. Expansion of covered services should include a diverse range of mental health professionals and modalities, catering to the unique needs of older individuals with mental and physical limitations. Cross-state licensure will enhance access to care via telemedicine, particularly in underserved areas. Continuous evaluation and iteration of ethics policies and applicable laws should occur in tandem with advancements in technology.

This evolving landscape requires a paradigm shift in medical training, with the incorporation of technology increasingly emerging as a core competency. A new skillset will be called for that includes a working understanding of several elements of data science. As we stand at the cusp of a future where technology will be omnipresent in healthcare, perhaps the question to be asked is not whether technology is a friend or foe but are we equipping the workforce to understand these tools, optimize their potential, and minimize the risks and drawbacks.

Sources of Support: An unrestricted gift from the Eric Warren Goldman Charitable Trust and the McLean Technology and Aging Lab.

Disclosures of Potential Conflicts of Interest: Dr. Vahia has performed scientific consultation for Otsuka. This work is not related to the current manuscript. He receives an editorial honorarium from the American Journal of Geriatric Psychiatry.

Other Sources of Funding: Dr. Vahia receives current research support from the National Institute on Aging, the National Institute of Mental Health, the Once Upon a Time Foundation, and the A2 Collective

Saeed Hashem, MD, is an academic geropsychiatrist at McLean Hospital and a clinical instructor of Psychiatry at Harvard Medical School in Boston. Ipsit V. Vahia, MD, is interim chief in the Division of Psychiatry, director, Digital Psychiatry Translation, and director, Technology and Aging Laboratory at McLean Hospital, and assistant professor of Psychiatry at Harvard. He may be contacted at

Photo credit: Frau aus UA



Aggarwal, R., Farag, S., Martin, G., Ashrafian, H., & Darzi, A. (2021). Patient perceptions on data sharing and applying artificial intelligence to health care data: cross-sectional survey. Journal of Medical Internet Research, 23(8), e26162.

Au-Yeung, W. M., Miller, L., Beattie, Z., May, R., Cray, H. V., Kabelac, Z., Katabi, D., Kaye, J., & Vahia, I. V. (2022). Monitoring Behaviors of Patients With Late-Stage Dementia Using Passive Environmental Sensing Approaches: A Case Series. The American journal of geriatric psychiatry: Official Journal of the American Association for Geriatric Psychiatry, 30(1), 1–11.

Bartels, S. J., DiMilia, P. R., Fortuna, K. L., & Naslund, J. A. (2018). Integrated care for older adults with serious mental illness and medical comorbidity. Psychiatric Clinics of North America, 41(1), 153–164.

Berridge, C., Turner, N. R., Liu, L., Fredriksen-Goldsen, K. I., Lyons, K. S., Demiris, G., Kaye, J., & Lober, W. B. (2023). Preliminary Efficacy of Let's Talk Tech: Technology Use Planning for Dementia Care Dyads. Innovation in Aging, 7(3), igad018.

Chiu, C. J., Hu, J., Lo, Y., & Chang, E. Y. (2020). Health promotion and disease prevention interventions for the elderly: a scoping review from 2015–2019. International Journal of Environmental Research and Public Health, 17(15), 5335.

Crigger, E., Reinbold, K., Hanson, C., Kao, A., Blake, K., & Irons, M. (2022). Trustworthy augmented intelligence in health care. Journal of Medical Systems, 46(2).

Dao, K. P., Cocker, K. D., Tong, H. L., Kocaballı, A. B., Chow, C. K., & Laranjo, L. (2021). Smartphone-delivered ecological momentary interventions based on ecological momentary assessments to promote health behaviors: systematic review and adapted checklist for reporting ecological momentary assessment and intervention studies. JMIR mHealth and uHealth, 9(11), e22890.

Jonk, Y., Burgess, A., Williamson, M. E., Thayer, D., Mackenzie, J., McGuire, C., Fox, K., & Coburn, A. F. (2020). Telehealth use in a rural state: a mixed‐methods study using maine's all‐payer claims database. The Journal of Rural Health, 37(4), 769-779.

Kuang, W., Zeng, G., Nie, Y., Cai, Y., Li, J., Yang, W., & Qiu, P. (2021). Equity in telemedicine for older adults during the COVID-19 pandemic. International Health, 14(3), 329–331.

Lieneck, C., Garvey, J. L., Collins, C., Graham, D. J., Loving, C., & Pearson, R. (2020). Rapid telehealth implementation during the COVID-19 global pandemic: a rapid review. Healthcare, 8(4), 517.

Mazziotti, R. and Rutigliano, G. (2021). Tele–mental health for reaching out to patients in a time of pandemic: provider survey and meta-analysis of patient satisfaction. JMIR Mental Health, 8(7), e26187.

Milne-Ives, M., Selby, E., Inkster, B., Lam, C., & Meinert, E. (2022). Artificial intelligence and machine learning in mobile apps for mental health: A scoping review. PLOS digital health, 1(8), e0000079.

Pywell, J., Vijaykumar, S., Dodd, A., & Coventry, L. (2020). Barriers to older adults’ uptake of mobile-based mental health interventions. Digital Health, 6, 205520762090542.

Schriger, S. H., Klein, M. R., Last, B. S., Fernandez-Marcote, S., Dallard, N., Jones, B., & Beidas, R. S. (2022). Community mental health clinicians’ perspectives on telehealth during the covid-19 pandemic: mixed methods study. JMIR Pediatrics and Parenting, 5(1), e29250.

Similä, H., Immonen, M., Toska-Tervola, J., Enwald, H., Keränen, N., Kangas, M., Jämsä, R., & Korpelainen, R. (2018). Feasibility of mobile mental wellness training for older adults. Geriatric Nursing, 39(5), 499–505.

Singh, S., & Germine, L. (2021). Technology meets tradition: a hybrid model for implementing digital tools in neuropsychology. International review of psychiatry (Abingdon, England), 33(4), 382–393.

Skurla, M., Rahman, A., Salcone, S., Mathias, L., Shah, B., Forester, B., & Vahia, I. (2022). Virtual reality and mental health in older adults: A systematic review. International Psychogeriatrics, 34(2), 143–155.

Tabibzadeh, M., and Tran, H. (2022). Improving quality of care and patient safety in virtual visits: investigating barriers of telehealth implementation. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 66(1), 1564–1568.

Vahia, I. V., Dickinson, R. A., & Trueba, A. F. (2022). Are Mobile Apps in Geriatric Mental Health Worth the Effort? The American Journal of Geriatric Psychiatry, 30(5), 585–587.

Vijayan, V., Connolly, J. P., Condell, J., McKelvey, N., & Gardiner, P. (2021). Review of wearable devices and data collection considerations for connected health. Sensors, 21(16), 5589.

Wilson, S. A., Byrne, P., Rodgers, S. E., & Maden, M. (2022). A Systematic Review of Smartphone and Tablet Use by Older Adults With and Without Cognitive Impairment. Innovation in aging, 6(2), igac002.

World Health Organization. (‎2023)‎. Regulatory considerations on artificial intelligence for health.

Ye, J. and Liu, A. (2022). Psychological consultation and health analysis method for artificial intelligence multidecision support. Security and Communication Networks, 2022, 1–9.

Zhang, G., Vahia, I. V., Liu, Y., Yang, Y., May, R., Cray, H. V., McGrory, W., & Katabi, D. (2021). Contactless In-Home Monitoring of the Long-Term Respiratory and Behavioral Phenotypes in Older Adults With COVID-19: A Case Series. Frontiers in psychiatry, 12, 754169.