The key facts

  • Nuclear conflict and malicious AI use are placed in the highest category of global threats in a Lancet assessment covering 2025–2100 and 204 countries and territories.
  • Experts identified 17 major risks, including climate change, pandemics, antibiotic resistance, and armed conflicts.
  • The probability of nuclear war and malicious AI use remains low, but their potential consequences are extremely large.

Nuclear conflict and the malicious use of artificial intelligence could pose a threat to the existence of humanity by the end of the 21st century.

They have been placed in the highest category of global threats in an assessment published in the journal The Lancet.

A commission of experts examined hundreds of potential threats and identified the 17 most significant risks to human health and survival for the period from 2025 to 2100. The study covers 204 countries and territories.

Nuclear war is dangerous not only because of its immediate destruction and human casualties, but also because of possible global famine caused by nuclear winter. The malicious use of AI, in turn, could facilitate the development of biological weapons.

Other serious threats include climate change, air pollution, antibiotic resistance, pandemics, conventional armed conflicts, child malnutrition, obesity, smoking, excessive alcohol consumption, and mental disorders.

The researchers assessed the potential consequences of these threats in terms of the loss of years of healthy life. In their view, timely measures in healthcare, technology, security, and public policy could significantly reduce the scale of future damage.

At the same time, the authors of the assessment emphasize that the probability of nuclear conflict and the malicious use of AI remains low, although the potential consequences of these scenarios are extremely large.