AI provides a timely example of the difficulty of translating innovation into better health. Its applications are expanding rapidly in medical imaging, clinical decision support, drug discovery, health communication, and disease surveillance. Substantial investment and technological progress have generated both optimism and uncertainty among patients and the public. The central question has become whether its use improves health outcomes, strengthens healthcare systems, and reduces inequalities.

Patients’ relationship with technology is also changing. Health professionals have traditionally selected medical technologies, with patients largely relying on those decisions. Increasingly, however, individuals consult generative AI directly to interpret symptoms, obtain information about medications, and reconsider clinical advice. This shift presents two distinct governance challenges: ensuring the safety and effectiveness of AI used in professional healthcare services and protecting people who independently use AI-based health information outside regulated clinical settings. Addressing both challenges requires reliable information, appropriate safeguards, and clear accountability.

Even impressive technical performance should not be equated with clinical effectiveness or benefits to population health. Gains in diagnostic accuracy, workflow efficiency, or analytical capability do not necessarily reduce mortality, improve quality of life, or increase health equity. Recent discussions in The Lancet Public Health have stressed the need to evaluate AI in relation to the broader functions and responsibilities of public health systems [2]. This evaluation should encompass real-world effectiveness, cost-effectiveness, patient experience, and the distribution of benefits and risks across population groups.

Public policy should therefore begin with the health problem to be addressed rather than with the technology. AI may help address shortages of health professionals, improve continuity of care, and strengthen disease surveillance. Introducing sophisticated technologies may nevertheless yield limited benefits if adequate personnel, reliable data, appropriate financing, and effective service delivery systems are lacking. Adoption driven by technology could even reinforce existing inequalities if it disproportionately benefits populations with better digital access, health literacy, and healthcare resources.

International frameworks offer guidance on these questions. In 2021, the World Health Organization (WHO) established six principles for the ethical governance of AI, emphasizing human autonomy, well-being and safety, transparency, accountability, inclusiveness and equity, and sustainability [3]. WHO subsequently issued guidance on the specific risks of large multimodal models [4]. The FUTURE-AI framework proposed practical standards for trustworthy AI across development, validation, deployment, and monitoring [5]. Putting these principles into practice through independent evaluation, effective regulation, and public accountability remains a substantial challenge.

In the Republic of Korea, technological innovation should form part of broader health system reform rather than be pursued as an isolated objective. Decisions about public investment, health insurance reimbursement, and institutional incentives should reflect demonstrated improvements in clinical effectiveness, population health, equity, and efficiency. Health professionals need appropriate training, and patients and communities should have a meaningful role in decisions affecting their care. These priorities are particularly important in underserved regions, where technology alone cannot resolve weaknesses in healthcare delivery.

Medical AI raises a question that applies more broadly to public health research: how can scientific advances and increasingly sophisticated analytical methods improve population health? Genomic technologies, mathematical models, environmental monitoring, and population-based health data can help us understand health risks and disease patterns. Methodological sophistication, however, is not sufficient to establish scientific or public health value. Reliable evidence, appropriate study design, reproducibility, and meaningful interpretation remain essential. The growing use of AI in research also reinforces the responsibility of researchers and journals to ensure transparency, verification, and scientific accountability.

The articles in this issue of Osong Public Health and Research Perspectives show how scientific evidence can help address a range of public health challenges. The perspective on surrogacy regulation in Africa highlights ethical safeguards, human rights, and institutional accountability as priorities for reproductive healthcare. A study of indoor exposure to volatile organic compounds examines associations between environmental measurements and urinary biomarkers, demonstrating the value of combining environmental monitoring with population-based health information. Mathematical modeling of Wolbachia-based dengue control explores the potential of quantitative approaches to inform disease prevention strategies. A study comparing welfare facilities for people with disabilities by coronavirus disease 2019 outbreak status emphasizes the need to prevent infection among vulnerable populations. Using national health screening statistics from 2014 to 2023, a study of metabolic syndrome components identifies changes in cardiometabolic risk patterns across age and sex groups and provides evidence for targeted prevention. Finally, a report on a foodborne outbreak associated with a festival demonstrates the practical contribution of genomic epidemiology to outbreak investigation and food safety surveillance.

Despite differences in methods and scope, these articles share the objective of generating reliable evidence to inform effective public health action. Their significance should be judged by the validity of their findings, their relevance to population health needs, and their potential to improve health outcomes, rather than by methodological sophistication or data volume alone.

The challenge is therefore to do more than accelerate scientific and technological innovation: it is to establish the conditions in which innovation produces meaningful and equitable health benefits. Meeting this challenge requires linking research to public health priorities, strengthening institutional capacity, evaluating real-world outcomes, and maintaining public accountability.

Innovation should be judged by what it achieves for population health, not merely by what technology can do. Scientific progress becomes meaningful when trustworthy evidence is put into practice to achieve effective prevention, accessible healthcare, reduced inequalities, and better health for all.

Citations to this article as recorded by