The World Health Organization (WHO) has called for stronger ethics review and oversight of artificial intelligence (AI)-related health research. In a new report titled Artificial Intelligence-related Health Research: Ethics Review and Oversight, released and launched around 21 September 2026, the agency warns that existing systems may not fully address novel risks linked to data, bias, fairness, accountability, privacy and the rapid deployment of AI tools.
AI is rapidly transforming how health data are analysed, how research is conducted and how new technologies are developed. While this creates opportunities to accelerate scientific discovery and improve health outcomes, the WHO stresses that without robust safeguards, AI-related research could undermine human rights, equity and public trust.
Three Categories of AI-Related Health Research
The report examines ethics oversight across three broad categories:
Health-related data science using AI
Research conducted with AI tools and technologies
Health-related research on AI tools and technologies themselves
Each category raises distinct ethical questions that traditional research ethics frameworks may not fully cover.
Key Challenges Highlighted
WHO identifies several critical issues:
Bias and fairness – Algorithms trained primarily on data from high-income settings can perform poorly or cause harm when applied to underrepresented populations.Transparency and accountability – Lack of clarity about how AI systems reach decisions makes it harder to assign responsibility when things go wrong.
Privacy and data governance – Even the use of anonymised or publicly available data can raise ethical concerns, including risks of re-identification or misuse.
Rapid deployment risks – AI-enabled tools may move from research into real-world use faster than oversight mechanisms can adapt.
Inequities for low- and middle-income countries – Concerns include data colonialism, ethics dumping, limited benefit sharing and power imbalances between researchers in different income settings.
Research ethics committees (RECs) remain central to protecting participants, but the report notes they often need additional expertise, training and resources to evaluate complex AI studies. Some AI research may fall outside traditional ethics review because it does not directly involve human participants, yet ethical risks can still exist.