Lancet assessment flags AI misuse among major health threats

A global health assessment published in The Lancet identifies 17 catastrophic risks that could devastate human health by 2100, with deliberate AI misuse and nuclear war singled out as the most extreme threats. The assessment warns that the gravest dangers may come not from a single crisis, but from overlapping pressures including technological advances, geopolitical tensions, climate change, inequality and preventable disease.

Researchers examined data from over 200 countries and drew on expertise across health, economics, national security, governance and technology. The assessment does not predict collapse, but argues that risks remain manageable if governments act early. AI could accelerate bioweapon design, cyberattacks on power grids or coordinated disinformation, while a large nuclear exchange could trigger wider food disruption through soot blocking sunlight and harming agriculture.

Other risks include climate change, air pollution, infectious disease, antimicrobial resistance, obesity, high blood pressure, smoking, heavy drinking, conventional wars, declining international health aid, hunger, poverty, limited education access, mental health burdens and unhealthy ageing. The commission calls for stronger health systems, emergency planning, scientific investment, early warning systems and coordinated action across health, security, education and environmental policy.