Lancet commission puts AI and nuclear war on an extinction-risk list

Measuring threats in healthy years

The commission used disability-adjusted life years, or DALYs, to estimate how many years of healthy life humanity could lose to disease, disability and premature death under current trends. To make its list, a threat had to carry the potential for large losses and have possible ways to address it.

The 17 threats identified were:

The report says nuclear conflict could trigger nuclear famine if particles thrown into the upper atmosphere blocked sunlight. It also warns that AI could be used to create biological weapons. In an accompanying editorial, The Lancet describes the realization of either scenario as apocalyptic.

Prevention changes the numbers

The commission modeled best- and worst-case outcomes for each threat. Limiting global warming to about 1.8 °C (3.2 °F), rather than allowing it to reach 4.4 °C, could prevent the loss of 1.89 billion healthy years.

The leading risks also differ by income level. In high-income regions, noncommunicable diseases such as obesity and mental health problems account for major threats. In low-income regions, health losses are more closely tied to factors such as low education levels.

International health aid is another point of exposure. Over the past year, major donor countries including the US and UK announced significant spending cuts. The commission estimates that ending or sharply reducing international health aid could cost up to 1.5 billion healthy years by 2100. Nearly 70% of those losses would fall on sub-Saharan Africa, which already bears the greatest burden of childhood malnutrition.

Ibrahim Abubakar, a commission member and professor of infectious disease epidemiology at University College London, said the assessment is meant to help national and international decision-makers understand the risks they face and how much action can change outcomes—not to frighten people with a long list.

A wider brief for governments

The commission has 35 experts, and its findings are due to be presented Sunday at the World Health Summit in Berlin, Germany. The authors plan to update the assessment annually, revising the 17-threat list as new data and risks emerge.

They urge leaders to invest in technology development and strengthen health systems before threats become crises. Co-chair Christopher Murray, director of the University of Washington’s Institute for Health Metrics and Evaluation, said the work began three years ago with a focus on conflict, climate and pandemics. The risks now appear broader, he said, and governments need to assign funding and staff to anticipate, prevent and prepare for them.

I think the report’s most useful claim is also its hardest test: that a broad risk register can guide action rather than simply expand the list of things governments are asked to worry about. Its annual updates will matter only if they show which interventions change the projected losses—and who is responsible for making them happen.