Artificial intelligence in medicine has advanced rapidly, but a new Japanese study suggests that growing familiarity with the technology has not translated into greater willingness to let it lead patient care. Researchers repeated the same online survey in 2018 and 2025 and compared responses from physicians and members of the general public, finding that intentions to use relatively autonomous medical AI were essentially unchanged across the seven-year gap.
The study included 399 physicians and 600 public respondents in 2018, and 300 physicians and 300 public respondents in 2025. Because the identical questionnaire was used in both waves, the design allowed a direct repeated cross-sectional comparison of attitudes in the two groups.
Mean responses to the key item assessing intention to use medical AI did not differ significantly between survey years for either physicians or the public. This stasis stood in contrast to consistently favorable evaluations of the technology’s usefulness and its future necessity, suggesting that positive appraisals alone do not drive acceptance of AI-led care.
Among physicians, the unadjusted mean response on medical AI’s potential to improve clinical efficiency was higher in 2025 than in 2018. However, this difference was no longer significant after adjustment for age and sex, indicating that demographic differences between the physician samples may account for the apparent change.
The public showed a different pattern. Self-rated knowledge of medical AI was higher in 2025, and expectations that the technology would increase healthcare costs were also greater. The authors conclude that increased self-rated knowledge may not necessarily be accompanied by greater intention to use relatively autonomous medical AI.
The findings point to specific concerns that implementation strategies should address alongside efforts to improve understanding: responsibility for AI-informed decisions, safety, data protection, and healthcare costs.
As an online survey conducted in Japan with self-identified participants, the study reflects attitudes in that setting at the two time points, and the reported differences are limited to the comparisons described.
Subject of Research: Physician and public acceptance of autonomous medical AI in Japan over time
Article Title: Changes in the acceptance of artificial intelligence in medicine among physicians and the general public in Japan: A repeated cross-sectional comparison of 2018 and 2025
Article References: Suzuki, T., Tamori, H., Ohashi, K., Morii, Y., Hirata, K., Kudo, K., & Ogasawara, K. (2026). Changes in the acceptance of artificial intelligence in medicine among physicians and the general public in Japan: A repeated cross-sectional comparison of 2018 and 2025. PLOS Digital Health, 5(10), e0001769. https://doi.org/10.1371/journal.pdig.0001769
Image Credits: AI Generated
DOI: 10.1371/journal.pdig.0001769
Keywords: medical AI, Japan, physicians, public attitudes, cross-sectional survey, AI acceptance, healthcare costs, clinical efficiency, data protection, survey research, Changes, acceptance
News Source: Ophelia Keating. (October 23, 2026). Seven Years On, Japanese Doctors and Public Still Hesitant About AI-Led Care. Scienmag.
