On a lesser scale, obesity would cause the greatest avoidable harm, while low education, inequality, and warming are also serious threats.

A global commission convened by The Lancet journal has identified 17 “interconnected catastrophic threats” to human health through 2100, singling out nuclear conflict and the malicious use of AI as the two risks that could potentially drive humanity to extinction.

The Lancet Commission on 21st Century Global Threats to Health examined 204 countries and territories, and estimated how many years of healthy life each threat could cost by the end of the century, according to CNN.

The commission’s central message is that leaders still have time to act. “The striking finding is not only the scale of these threats, but that humanity still has the agency to act,” said co-chair Dr Christopher Murray, director of the University of Washington’s Institute for Health Metrics and Evaluation.

Chosen methodology

The commission used Global Burden of Disease data to estimate avoidable harm.

  • Obesity could account for 23.1bn years of healthy life lost by 2100, and policies that lower obesity risk could prevent that loss
  • Low educational attainment, especially in sub-Saharan Africa, could cost 16.1bn healthy years
  • Income inequality could cost 13.8bn
  • Keeping global warming to roughly 1.8°C rather than 4.4°C could avert 1.89bn lost healthy years
  • The remaining threats include pandemics, antimicrobial resistance, poor mental health, high blood pressure, smoking, childhood malnutrition, unhealthy aging, air pollution, climate change, conventional conflict, declining development assistance for health, high alcohol use, and other risks
  • Reducing the danger of malicious AI could prevent about 10.4bn lost healthy years
  • Preventing nuclear conflict could avert an estimated 133million deaths

Murray has cautioned that, because these events lack historical precedent, “these estimates carry much greater uncertainty than those for other threats.” The commission has examined three ways AI could cause harm: helping people develop biological weapons; system failures such as autonomous weapons; and loss of human control over advanced systems.

An accompanying editorial said the report is not meant to create a fixed ranking but to “catalyze international, regional, and national actions.” Lancet editor-in-chief Richard Horton said: “The jeopardies that face us are multiple. Our solutions need to match that reality.” Said co-chair Dr Natalia Kanem, former head of the UN Population Fund: “Recent global events show us just how quickly stability can give way to crisis, and how costly it is to respond only after disaster strikes. Yet the future is not inevitable.

The findings were released ahead of the World Health Summit in Berlin. Murray is scheduled to lead a virtual media workshop on the findings on 12 October 2026.