The World Health Organization (WHO) officially released a new report titled "Ethical Review and Regulation of AI-Related Health Research" on September 21, 2026. The report states that while artificial intelligence is reshaping health research at an unprecedented speed, accelerating scientific discovery, and improving health outcomes, AI health research could also introduce new risks without strong ethical safeguards and regulatory oversight, thereby endangering human rights, social equity, and public trust.

Regarding existing ethical review mechanisms, the WHO warns that traditional systems often struggle to address the new risks brought by artificial intelligence, including algorithmic transparency, bias, fairness, accountability, privacy protection, and potential harms from rapidly deploying AI tools. To this end, the report clearly categorizes AI-related health research into three main types: research involving health-related data using AI; research conducted using AI tools and technologies; and health-related research specifically focused on AI tools and technologies themselves. Additionally, the report explores how to strengthen ethical oversight throughout the entire lifecycle of research, from design, ethical review, publication, regulation, to implementation.

With regard to the challenges and opportunities faced by low- and middle-income countries, the report emphasizes that although artificial intelligence holds great potential in addressing long-standing medical challenges, the majority of current AI research and technological development remains concentrated in high-income regions. Therefore, strengthening local leadership, capacity building, and broad participation in AI health research will be key to ensuring fair sharing of benefits and avoiding new forms of exclusion or inequality.

The WHO emphasizes that Research Ethics Committees (RECs) will continue to play a central role in protecting research participants, but they need additional expertise, training, and resources to effectively address increasingly complex AI-related research. Moreover, multiple stakeholders, including funders, scientific journals, data governance bodies, professional associations, and regulatory agencies, must work closely across disciplines and sectors to ensure that AI medical innovations remain safe, trustworthy, and beneficial to all humanity.