Artificial intelligence (AI) is one of the world’s most resource-intensive digital technologies.1 Nonetheless, the major concerns raised by the environmental impact of AI on health have not been sufficiently addressed in the fields of global health or bioethics. AI ethics frameworks have, to date, primarily focused on the core concerns of transparency, justice and fairness, non-maleficence, responsibility, and privacy, whereas the environmental impact of AI has been relegated to a subsidiary consideration. When included, the environmental impact of AI often falls under the somewhat under-represented ethical concern of sustainability—it is rarely considered as a key ethical concern in its own right.2 Given the established health burden of climate change, in this Viewpoint, we point to a central, unaddressed tension: the AI solutions proposed to address global health concerns are exacerbating the very health problems they purport to address. AI cannot be a force for good in global health until the adverse effects of AI on climate and health are addressed. Centring the environmental impact of AI on health in the agendas of both global health and the bioethics and technology sector requires refiguring the core principles of AI ethics frameworks and fostering proactive approaches for AI that prioritise reducing environmental impact.
