JMIR Aging. 2026 Apr 21;9:e85626. doi: 10.2196/85626.
ABSTRACT
BACKGROUND: With the growing aging population, technology that supports independent living is increasingly important. Web search systems are well established, whereas generative artificial intelligence (Gen-AI; eg, ChatGPT) represents a newer, adaptive tool that could offer personalized information access. However, little is known about how older adults, particularly those with mild cognitive impairment (MCI) or mild dementia, perceive and engage with these systems.
OBJECTIVE: This study explored the use of and perspectives on web search and Gen-AI in older adults with and without cognitive impairment (including MCI and early-stage dementia).
METHODS: A UK-wide mixed methods study was conducted with older adults, including those with MCI or mild dementia. An online survey captured technology use, Likert-scale ratings of web search and Gen-AI, and reasons for nonuse. Follow-up focus groups provided in-depth qualitative perspectives. Quantitative data were analyzed using descriptive and comparative statistics, while qualitative data were thematically analyzed.
RESULTS: Survey findings showed higher use of web search (275/280, 98.2%) compared to Gen-AI (40/286, 14%) within these groups. Web search was rated positively across participants, although challenges were raised regarding the phrasing of queries and commercialization. Gen-AI use was less common, but more than half of nonusers expressed willingness to adopt it in the future. Combined with focus group responses, themes exploring keyword searching, mistrust, lack of knowledge, and willingness to learn were established. Participants also suggested potential applications of Gen-AI, such as supporting independent living through monitoring and simplifying complex searches.
CONCLUSIONS: Web search remains the primary method, and participants highlighted both advantages and frustrations with current systems. Gen-AI was underused but seen as promising, with its adoption mainly limited by mistrust and knowledge gaps. Our findings indicate that structured training, early introduction, and user-centered design could encourage adoption, enhance accessibility, and support independent living among older adults with and without MCI.
PMID:42013397 | DOI:10.2196/85626
