GenAI Acceptance in Dementia Caregivers Splits Along Race and Age Lines
As generative AI moves closer to routine health care use, this survey points to a familiar equity problem in a new form: the people most likely to benefit from better caregiver communication may not see the technology in the same way. Investigators surveyed 527 family members of a person with Alzheimer’s Disease and Related Dementia, or memory issues, and found clear differences in how respondents weighed the risks and benefits of GenAI for caregiving support.
The study was designed around a practical clinical question. Can automated language tools help family members who carry much of the day-to-day burden of dementia care without deepening mistrust or widening existing disparities? The answer was not uniform across the sample. Baseline attitudes depended on race, age, and caregiving context, and those differences matter because GenAI is most likely to be deployed first in exactly these settings: reminders, educational prompts, and tailored messages meant to help relatives navigate an often confusing care journey.
Black Participants Expressed More Favorable Views of AI Tools
A striking racial difference emerged in the analysis. Black or African American participants were more than twice as likely as White participants to express an overall positive view of the AI tools, with an odds ratio of 2.23 and a 95% confidence interval of 1.43 to 3.48; the association was highly significant at p < 0.001. In a field where adoption is often assumed to be constrained by distrust, that pattern complicates the usual narrative.
For dementia caregiving, the result may reflect a practical calculation. Families are often asked to absorb complicated symptom changes, medication changes, and safety decisions while juggling work and other responsibilities. A system that can generate clear, personalized, and timely messages could reduce friction in those exchanges, especially when care coordination is fragmented. GenAI may also appeal when it offers information in a more accessible format than standard health system messaging, which is often dense, delayed, or poorly tailored to the needs of family caregivers.
One sentence stands out from the data: enthusiasm was not evenly distributed, but neither was skepticism.
Younger Adults Were More Cautious Despite Recognizing the Benefits
Age shaped the response profile in a different way. The benefits of GenAI were recognized well across age groups, but adults aged 25–34 were more cautious than older respondents. That finding is clinically important because younger family members often sit at the intersection of caregiving logistics, digital literacy, and household technology use. They may be the very relatives who manage appointment portals, messaging systems, and care coordination, yet they appear more likely to interrogate how AI-generated content could fail.
Their caution is easy to understand. GenAI systems can produce polished text without offering transparent reasoning, and in dementia care an inaccurate reassurance or an overconfident recommendation can have real consequences. A tool that drafts a message about wandering risk, medication adherence, or behavioral change has to be trusted not just for readability but for factual restraint. For a younger adult who already knows how quickly digital tools can overpromise, hesitation may represent a form of quality control rather than resistance.
Why the Socioecological Frame Matters for Dementia Care
The investigators examined attitudes across three representative scenarios at the individual level of the socioecological model. That choice is more than a theoretical flourish. Dementia caregiving rarely happens in isolation; it is shaped by the family member’s own workload, digital access, trust in institutions, and prior exposure to health technology. A message-generating system that appears efficient in a lab setting can fail in the home if it ignores those layers.
Viewed through that lens, the findings suggest a useful implementation target. GenAI adoption should not be built around a generic user. It should be co-designed for caregivers who bring different expectations about privacy, personalization, and clinical authority. Community partners, especially younger relatives, can identify where automated communication feels helpful and where it feels intrusive, misleading, or simply too blunt for a sensitive family situation. If those perspectives are ignored, a tool meant to support high-risk families could reproduce the same inequities health systems are trying to eliminate.
Equity Will Depend on Who Shapes the Tool Before It Scales
The study’s clinical relevance lies in implementation, not hype. GenAI has the potential to ease the burden on health professionals and improve care coordination, but the path to that goal runs through trust, interpretation, and cultural fit. In dementia care, where family members are often the first to notice change and the first to absorb uncertainty, a messaging system must be acceptable before it can be useful.
This survey also points to a broader design challenge: health systems cannot assume that enthusiasm among one group will translate to universal adoption. Responses from Black participants suggest substantial openness to the technology, while the caution among younger adults shows that perceived benefit does not eliminate concern. Those differences should shape how clinicians, researchers, and developers frame GenAI use in caregiving workflows.
The study was cross-sectional and limited to an online sample of family members recruited through social media and flyer postings at an urban medical center, so it captures attitudes at a single point in time rather than behavior after real-world deployment. Even so, the signal is clear enough to guide the next phase of work: test GenAI with the people who will use it, not just around them.