Crossing the Digital Divide: Understanding Older Adults' Technology Acceptance of Generative Artificial Intelligence:...

The diffusion of generative artificial intelligence is redefining the relationship between older adults and digital technologies. Existing studies have largely explained the digital divide among older adults through traditional digital media such as smartphones and mobile payment, while insufficient attention has been paid to how conversational artificial intelligence changes older adults' entry thresholds, sustained-use motivations, and risk-judgment mechanisms. This study adopts a multimethod qualitative design, with semistructured interviews as the primary method and digital scenario-based tasks as a complementary method. From January to February 2026, the study was conducted in Chedun Town and Xinqiao Town, Songjiang District, Shanghai, with 40 participants aged 60 and above who had internet-use experience and showed a pronounced tendency toward high-frequency smartphone use. The data were analyzed through thematic analysis. Interview transcripts, task-observation records, screenshots of AI conversations, and field notes were triangulated; an initial coding manual, double review, and consensus discussion were used to enhance analytic credibility. The study finds that generative artificial intelligence shifts the technological threshold from procedural operation to need articulation and result evaluation, enabling some older adults to compensate for limitations in traditional interface operation through life experience and language-expression skills. Older adults' sustained use is not driven simply by efficiency, but is closely related to emotional anchoring in later life, experience organization, relationship maintenance, and concrete task handling. Community support, peer learning, and intergenerational assistance constitute important mediating mechanisms through which older adults move from initial experimentation to stable use. Whether older adults truly cross the digital divide depends on whether they can develop the capacities of source verification, contextual delimitation, and boundary judgment. Based on these findings, the study proposes a "conditional crossing" mechanism for older adults' acceptance of AI technologies and argues that age-friendly governance should move beyond interface simplification toward capability enhancement and risk support.