A new global risk assessment published in The Lancet warns that the malicious use of artificial intelligence and nuclear war represent the most severe risks to humanity's future by 2100. The evaluation places these two catastrophic hazards at the peak of a pyramid of 17 major threats to human health across the coming decades.
The research was conducted by the Lancet Commission on 21st-century global threats to health, which includes 37 specialists in health, security, economics, and technology. The team presented its conclusions at the World Health Summit in Berlin, Germany.
To measure and contrast risks, the authors utilized disability-adjusted life years (DALYs) across 204 countries and territories from 2025 to 2100, broken down by age, sex, and cause. For comparative context, researchers noted that the COVID-19 pandemic generated roughly 500 million DALYs.
Beyond direct impacts, the assessment highlights cascading outcomes, noting that nuclear conflict encompasses widespread nuclear famine, while malicious AI includes the creation and deployment of bioweapons. The wider list of evaluated risks includes antimicrobial resistance, climate change, pandemics, income inequality, and health issues such as obesity, tobacco use, and mental health challenges.
The commission highlighted that while nuclear conflict and hostile AI present distinct extinction risks within the next 75 years, all 17 examined threats remain deeply interconnected. Consequently, the authors called for a unified response framework rooted in science, human rights, and global solidarity. How might international institutions better coordinate to manage these linked global hazards?