Key Points
- Despite significant technological advancements in artificial intelligence over the past seven years, intentions to adopt relatively autonomous medical AI remain virtually unchanged among both Japanese physicians and the general public.
- Both doctors and the public consistently evaluated medical AI as useful and necessary for the future, demonstrating that favorable perceptions do not automatically lead to a willingness to allow AI to take a leading role in clinical care.
- Researchers conducted a repeated cross-sectional online study comparing identical survey datasets collected in 2018 (399 physicians, 600 public) and 2025 (300 physicians, 300 public).
- An initial unadjusted increase in physicians’ belief that AI could improve clinical efficiency in 2025 became statistically non-significant once controlled for age and sex, suggesting sample demographic differences influenced the initial result.
- Members of the public reported higher self-rated knowledge of medical AI in 2025 compared to 2018, but they also expressed heightened expectations that the technology would drive up overall healthcare costs.
- The study highlights key implementation challenges that go beyond public education, specifically pointing to unresolved concerns surrounding legal responsibility, patient safety, data protection, and healthcare expenditure.
Tokyo (Japan Daily Sun) October 24, 2026 – Rapid technical progress in medical artificial intelligence over the past seven years has failed to increase the willingness of Japanese doctors or the public to adopt autonomous AI-led healthcare. A longitudinal repeated cross-sectional study comparing survey data from 2018 and 2025 shows that intentions to use autonomous clinical AI systems have remained stagnant across both demographic groups. Although respondents consistently acknowledged the usefulness and future necessity of AI, positive sentiment has not translated into active acceptance of autonomous systems. The findings demonstrate that increasing technical familiarity alone is insufficient to drive clinical integration without addressing fundamental concerns such as legal liability, safety, data privacy, and medical costs.
Why Has Seven Years of Technological Advancement Not Increased AI Acceptance in Healthcare?
As reported by science journalist Ophelia Keating of Scienmag, artificial intelligence tools in clinical settings have advanced dramatically in capability and scope over the last seven years, yet user willingness to rely on autonomous medical systems has shown no measurable increase.
To evaluate shifts in attitude, researchers conducted a repeated cross-sectional comparative study using identical questionnaire instruments administered at two distinct time periods: 2018 and 2025. The initial 2018 survey wave gathered responses from 399 physicians and 600 members of the general public. In the 2025 follow-up wave, the study surveyed 300 physicians and 300 members of the public.
Across both groups, the mean responses evaluating the key metric—the intention to use relatively autonomous medical AI systems—did not show a statistically significant difference between 2018 and 2025. As Ophelia Keating noted in Scienmag, this persistent stasis directly contrasted with the consistently high ratings both groups gave regarding the technology’s overall usefulness and long-term necessity in medicine. The authors highlighted that positive valuations of AI’s technical utility do not inherently drive adoption when it comes to autonomous patient care.
How Do the Views of Japanese Physicians Compare to the General Public?
As documented in the study details published by Scienmag, physicians and members of the general public exhibited distinct patterns in their responses regarding medical AI, though neither group demonstrated an increased intent to deploy autonomous care systems.
Among medical practitioners, the unadjusted mean scores regarding AI’s capacity to improve clinical efficiency appeared higher in the 2025 sample than in 2018. However, as reported by Ophelia Keating, when researchers adjusted the statistical model to account for age and sex, this apparent increase lost statistical significance. This statistical adjustment indicates that shifting demographic distributions within the physician cohorts, rather than a genuine overarching shift in professional opinion, likely accounted for the higher raw efficiency scores in 2025.
Conversely, the general public exhibited a distinct trend driven by self-perceived literacy and economic expectations. Public respondents in 2025 reported significantly higher levels of self-rated knowledge regarding medical AI than participants in 2018. Alongside this increased self-reported understanding, however, public expectations that medical AI integration would elevate overall healthcare costs also grew substantially. The study authors concluded that higher self-rated knowledge does not automatically translate into a greater willingness to utilize autonomous medical technology in practice.
Background of the Particular Development
The integration of artificial intelligence into medical care has been a strategic focus for healthcare systems worldwide, with Japan standing as a key focal point due to its rapidly aging population and shrinking medical workforce. Over the past decade, major investments have been directed toward diagnostic support tools, image recognition algorithms, and administrative automation to alleviate burdens on clinical staff.
Despite these technological rollouts, regulatory framework developments and public trust initiatives have often lagged behind technical capabilities. While lower-risk applications—such as AI-assisted diagnostic imaging or administrative sorting—have seen growing implementation, systems that operate with higher autonomy or make direct clinical decisions remain heavily scrutinised. The seven-year study tracking attitudes from 2018 to 2025 reflects a broader global debate over the boundaries of machine autonomy in healthcare, illustrating that while passive appreciation for technology has expanded, operational trust in doctorless or AI-led interventions has encountered a ceiling.
Prediction: How This Development Can Affect Healthcare Stakeholders and Patients
The persistent reluctance of both medical professionals and the public to embrace autonomous AI systems is expected to influence the trajectory of healthcare policy, technology development, and clinical operations in several distinct ways:
- For Healthcare Technology Developers: Technology companies and medical device manufacturers will likely shift their design strategies away from fully autonomous or “AI-led” systems toward human-in-the-loop, supportive tools. Products that emphasize physician oversight and decision augmentation rather than automated diagnosis will experience smoother regulatory approval and clinical uptake.
- For Policymakers and Regulators: Government agencies and legal bodies will be pressed to establish clearer liability frameworks. Until legislation clearly defines whether accountability for an AI error rests with the clinician, the health institution, or the software vendor, medical professionals will remain cautious about using autonomous features.
- For Health System Administrators: Hospital management will need to account for potential cost increases and patient hesitation when deploying AI systems. Implementation strategies will require transparent communication regarding data security, patient privacy, and financial impact to prevent patient resistance.
- For Patients and the Public: Patients are likely to see AI used predominantly behind the scenes in diagnostics and workflow efficiency rather than in direct, autonomous patient care. Human-centric care will remain the expected standard, with AI serving strictly as an auxiliary tool under explicit physician supervision.
