Edit Project

This study compared the performance of a generative artificial intelligence (AI) search tool, Microsoft Copilot, with a conventional search engine, Google Web Search, for assisting Australians in navigating health care information. Consumers at the 2024 Dalby Roundtable identified navigating the healthcare system as a particular challenge. Large language model (LLM)–powered tools may improve access by generating direct, conversational answers. Ninety-seven adults from Queensland, Australia, completed a web-based survey involving scenario-based health care navigation tasks. Participants used either Microsoft Copilot or Google Web Search, and their responses were evaluated for accuracy using binary, graded, and numerical scoring systems. User experience was assessed using the Technology Rating Questionnaire (TRQ). Microsoft Copilot outperformed Google Web Search on two of eight tasks—identifying aged care application services and listing mobility allowance eligibility criteria—but showed no clear advantage in the others. Participants rated Google Web Search higher for willingness to adopt, perceived impact on quality of life, and ease of learning, while both tools received similar ratings for confidence, perceived value, and privacy concerns. Overall, generative AI achieved comparable accuracy to traditional search engines but did not improve user experience. Further research is needed as AI tools evolve and users become more familiar with them. Citation: JMIR AI 2025;4:e76203. doi:10.2196/76203