How Social Robotics and Embodied AI Could Reshape Neurology Care

If you spend any time in a memory clinic or stroke unit, you can feel it in the room: neurologic disease rarely arrives alone. It brings isolation, late-night worry, and long stretches of time where nothing happens but waiting. We now know that loneliness is not just unfortunate background noise but a biologic insult. The U.S. Surgeon General’s office estimates that chronic loneliness and social isolation in older adults raise dementia risk by roughly 50%, on par with traditional vascular risk factors.¹ A Class I study in Neurology similarly found that loneliness significantly increases 10-year dementia risk.²

Neurology has embraced robotics on the motor side of the house. Robot-assisted gait training and exoskeletons are now part of mainstream poststroke rehabilitation, delivering high-intensity, repetitive practice that would be impossible for therapists alone.^3,4 But a quieter revolution is emerging around what we might call “social robots”, embodied systems that talk, listen, and respond, with the explicit goal of improving emotional and cognitive outcomes.

The evidence base is rapidly building. Pet-like robots such as PARO, the robotic baby seal, have been studied in randomized and cluster-randomized trials in people living with dementia. These studies report reductions in agitation and depression and signals of improved quality of life compared with usual care or non-interactive plush toys.⁵ Meta-analyses of social robots in older adults now show significant effects on depression and loneliness, with large effect sizes in some subgroups.⁶

Systems like ElliQ extend this model, combining a tabletop robot with proactive voice-based coaching, games, and wellness check-ins. Early trials suggest reductions in loneliness and improved engagement in health behaviors, and a growing number of state aging agencies now subsidize access for isolated older adults.⁷ The arc is clear: patients with neurological conditions– who face high burdens of disability, apathy, and caregiver strain – are an obvious next population to benefit.

Why Embodiment Matters

Most of us already interact with disembodied AI through smartphones and smart speakers. For neurologic populations, that may not be enough. Embodied, huggable companions engage a different set of neural and behavioral pathways: affective touch, multisensory integration, and procedural memory. For a patient with Alzheimer disease who no longer remembers a grandchild’s name but instinctively knows how to hug a teddy bear, a soft, weighty object that also talks back can be uniquely powerful.

That design philosophy is driving a new generation of devices, including Yaya Bear, a teddy bear with embedded voice AI that my team developed at CareYaya under a grant from the National Institute on Aging’s Startup Challenge.⁸ The bear requires no apps, passwords, or screens; older adults simply speak to it as they would to a grandchild or longtime friend. Over time, the system learns their life story, prompts reminiscence, and offers gentle conversation – while also generating structured insights about mood, engagement, and possible cognitive changes that can be shared (with consent) with families and clinicians.

In early pilots with people living with dementia, caregivers have reported decreased “sundowning” agitation, more spontaneous storytelling, and – most importantly – moments of joy they hadn’t seen in years.⁹ These observations echo quantitative work from Japan and Europe, where social robots have been shown to improve affective states and emotional responses in dementia care settings.¹⁰

Where Neurology Fits

For neurology clinicians, the question is not whether robots will enter the exam room, but rather, how to deploy them responsibly. An embodied AI companion is not a substitute for family, therapists, or a multidisciplinary team. But it can be a practical adjunct in three areas:

1. Dementia and behavioral symptoms. Social robots can provide structured daytime engagement, reduce behavioral and psychological symptoms of dementia, and offer families a way to “extend” their presence between visits.
2. Post-stroke and Parkinson disease mood disorders. Apathy, depression, and social withdrawal after stroke or in Parkinson disease are common and often under-treated. An always-available, non-judgmental companion can nudge patients to talk, move, and adhere to exercises or medications.
3. Remote monitoring and caregiver support. When combined with appropriate safeguards, conversational data can flag changes in speech fluency, content, or affect that may indicate delirium, progression, or emerging mental health concerns, prompting earlier intervention. (CareYaya’s QuikTok program backed by the National Institute on Aging uses a similar approach over the telephone.)¹¹

Ethical Guardrails and the Path Forward

The AMA Journal of Ethics has already asked whether AI should play a role in cultivating social connection among older adults, emphasizing transparency, consent, and the risk of deception.¹² Neurology professionals should lead here: clearly explaining to patients and families what these systems can and cannot do, avoiding anthropomorphism, and ensuring that commercial incentives do not drive over-deployment in vulnerable populations.

NeurologyLive^® readers are in a unique position to shape this field. The next step is not more glossy demos but rigor. We need pragmatic, multi-site trials that embed social robots like PARO, ElliQ, and Yaya Bear into memory clinics, movement-disorders centers, and stroke programs, with outcomes that matter: mood, function, caregiver burden, hospitalizations, and cost.

If loneliness and social isolation are now recognized as major neurologic risk multipliers, then it is not extravagant to experiment with new tools to address them. A soft, talking teddy bear will not cure dementia or rebuild a corticospinal tract. But if embodied AI can help a person with neurologic disease feel less alone, move a bit more, and stay safely at home a bit longer, it deserves a serious place in our clinical imagination and in our research agendas.

REFERENCES
1. U.S. Department of Health and Human Services. Social Connection. Accessed December 12, 2025. https://www.hhs.gov/surgeongeneral/reports-and-publications/connection/index.html
2. Salinas J, Beiser AS, Samra JK, et al. Association of Loneliness With 10-Year Dementia Risk and Early Markers of Vulnerability for Neurocognitive Decline. Neurology. 2022;98(13):e1337-e1348. doi:10.1212/WNL.0000000000200039
3. Warutkar V, Dadgal R, Mangulkar UR. Use of Robotics in Gait Rehabilitation Following Stroke: A Review. Cureus. 2022;14(11):e31075. Published 2022 Nov 4. doi:10.7759/cureus.31075
4. Lee JH, Kim G. Effectiveness of Robot-Assisted Gait Training in Stroke Rehabilitation: A Systematic Review and Meta-Analysis. J Clin Med. 2025;14(13):4809. Published 2025 Jul 7. doi:10.3390/jcm14134809
5. Moyle W, Jones CJ, Murfield JE, et al. Use of a Robotic Seal as a Therapeutic Tool to Improve Dementia Symptoms: A Cluster-Randomized Controlled Trial. J Am Med Dir Assoc. 2017;18(9):766-773. doi:10.1016/j.jamda.2017.03.018
6. Yen HY, Huang CW, Chiu HL, Jin G. The Effect of Social Robots on Depression and Loneliness for Older Residents in Long-Term Care Facilities: A Meta-Analysis of Randomized Controlled Trials. J Am Med Dir Assoc. 2024;25(6):104979. doi:10.1016/j.jamda.2024.02.017
7. Broadbent E, Loveys K, Ilan G, et al. ElliQ, an AI-Driven Social Robot to Alleviate Loneliness: Progress and Lessons Learned. JAR Life. 2024;13:22-28. Published 2024 Mar 5. doi:10.14283/jarlife.2024.2
8. National Institute on Aging. 2025 Start-Up Challenge Finalists and Mentors. Accessed December 12, 2025. https://www.nia.nih.gov/research/sbir/startup-challenge/2025-finalists-and-mentors
9. Shah NK. Guest column: Social robotics emerges as breakthrough in senior care. News. WRAL TechWire. November 24, 2025. Accessed December 12, 2025. https://www.wral.com/business/technology/social-robotics-breakthrough-senior-care-november-2025/
10. Otaka E, Osawa A, Kato K, et al. Positive Emotional Responses to Socially Assistive Robots in People With Dementia: Pilot Study. JMIR Aging. 2024;7:e52443. Published 2024 Apr 11. doi:10.2196/52443
11. CareYaya’s QuikTok is AI phone companion for lonely aging adults. News. CareYaya. Accessed December 12, 2025. https://www.careyaya.org/resources/news/quiktok
12. Portacolone E, Feddoes DE. Should Artificial Intelligence Play a Role in Cultivating Social Connections Among Older Adults?. AMA J Ethics. 2023;25(11):E818-E824. Published 2023 Nov 1. doi:10.1001/amajethics.2023.818

Neal K. Shah is a health care researcher specializing in caregiving, workforce innovation, and artificial intelligence. He is an NIH-funded Principal Investigator on the Johns Hopkins YayaGuide AI for Caregiver Training project and co-Principal Investigator of the University of Pennsylvania's Counterforce Health AI project to help patients and clinics appeal health insurance denials. Neal also serves on North Carolina's Steering Committee on Aging. He is CEO of CareYaya, Chairman of Counterforce Health and the author of Insured to Death: How Health Insurance Screws Over Americans - And How We Take It Back.