Two threats stand out for their potential to cause human extinction: nuclear conflict and the malicious use of AI. The assessment does not predict that either event will occur. Rather, it examines their potential consequences and the extent to which preventive action could reduce the risks.
What are the catastrophic threats?
The commission identified the following threats as major risks to global health and human survival between 2025 and 2100.
- Malicious use of AI: artificial intelligence could be misused to develop dangerous technologies, including biological weapons, or amplify other catastrophic risks. The commission places malicious AI use among the threats with potential to cause human extinction.
- Nuclear conflict: A nuclear war could cause mass casualties, disrupt food production and trigger widespread famine, with consequences extending far beyond the countries directly involved.
- Climate change: Rising temperatures, extreme heat and related environmental pressures threaten health and could intensify displacement, food insecurity and other global risks.
- Air pollution: Exposure to polluted air contributes to disease and premature deaths, with urban populations and vulnerable communities particularly exposed.
- Antimicrobial resistance: Drug-resistant infections threaten the effectiveness of antibiotics and other treatments, making routine infections harder to treat and medical procedures riskier.
- Childhood malnutrition: Inadequate nutrition can impair physical growth, brain development and immunity, with lasting consequences for health and human development.
- Conventional conflict: Wars can cause deaths and injuries, displace populations and destroy healthcare infrastructure, while interrupting food, water and medicine supplies.
- Declining development assistance for health: Cuts in international health funding can weaken healthcare systems, disrupt essential programmes and leave poorer countries less equipped to manage disease outbreaks.
- High alcohol use: Harmful alcohol consumption increases the risk of liver disease, certain cancers, cardiovascular problems, injuries and other health complications.
- High blood pressure: Uncontrolled hypertension raises the risk of strokes, heart attacks, kidney disease and premature death.
- Income inequality: Unequal access to income, nutrition, education and medical care can widen differences in life expectancy and leave disadvantaged populations more exposed to health shocks.
- Low educational attainment: Limited access to education can affect health literacy, employment prospects and access to healthcare, contributing to preventable illness and mortality.
- Naturally occurring pandemics: Emerging infectious diseases can spread across borders, overwhelm health systems and disrupt economies, as demonstrated by the COVID-19 pandemic.
- Obesity: Obesity increases the risk of diabetes, cardiovascular disease and several cancers, adding to the long-term burden on healthcare systems.
- Poor mental health: Mental health conditions can affect quality of life, social functioning and economic participation, while often remaining inadequately treated.
- Smoking: Tobacco use remains a major cause of preventable disease, including lung cancer, chronic respiratory illness and cardiovascular disease.
- Unhealthy ageing: Longer lifespans do not necessarily mean healthier lives. Age-related illness, disability and inadequate long-term care can increase the burden on families and health systems.
AI and nuclear war: Why the commission flags extinction-level risks
The inclusion of AI alongside nuclear conflict highlights how the definition of global health threats is expanding. The concern is not simply that AI could automate jobs or disrupt industries. Its malicious use could potentially amplify the development of dangerous biological agents, facilitate attacks on critical infrastructure or intensify other threats.
Nuclear conflict presents a more established pathway to mass destruction. Beyond immediate deaths, a large-scale nuclear exchange could disrupt agricultural production, damage essential infrastructure and expose populations far from the battlefield to severe consequences.
According to the commission's modelling, preventive action against nuclear conflict could avert an estimated 133 million deaths and 9.35 billion disability-adjusted life years (DALYs) by 2100. Reducing the risk of malicious AI use could avert an estimated 10.4 billion DALYs over the same period. These are modelled estimates of potential health losses avoided, not predictions of deaths or illnesses that will necessarily occur.
Climate change could cost billions of healthy life-years
Climate change is among the threats already imposing measurable harm, rather than a risk that exists only in a worst-case future.
The commission estimates that limiting global warming to around 1.8°C, instead of allowing temperatures to reach 4.4°C, could prevent 1.89 billion years of healthy life from being lost through impacts including extreme heat and air pollution.
The estimate captures direct health effects that can be measured, but does not fully account for harder-to-quantify consequences such as displacement, drought and mental health impacts. The actual burden could therefore be greater.
Climate change can also compound other threats. Drought and extreme weather can undermine food security, displacement can put pressure on urban infrastructure, and heat can worsen existing health conditions. The risks are consequently connected rather than independent.
What does DALY mean, why it matters?
The commission uses disability-adjusted life years, or DALYs, to assess the potential health burden of these threats. One DALY represents one year of healthy life lost because of premature death, illness or disability. The measure helps researchers compare risks that would otherwise be difficult to assess on the same scale.
For example, a nuclear conflict could cause large-scale premature deaths, while air pollution and high blood pressure may gradually contribute to millions of cases of illness and early death. DALYs provide a common framework for evaluating these different outcomes.
The commission identified a threat as catastrophic if it met three criteria:
- It could account for at least one billion DALYs by 2100.
- Sufficient data existed to model its potential impact.
- There was meaningful scope for policy intervention.
The exercise examined hundreds of possible threats before narrowing the assessment to 17 priorities. It modelled plausible best-case and worst-case futures to estimate the health losses that could be avoided through action.
Why poorer countries can be hit harder
Although the threats are global, their effects are unlikely to be distributed equally. Countries with stronger healthcare infrastructure, financial resources and emergency-response systems may be better equipped to manage outbreaks, environmental disasters and chronic diseases. In contrast, communities with limited access to healthcare, education, clean air and adequate nutrition may face greater harm.
The commission also stresses that risks can reinforce existing inequalities. A decline in international health assistance, for instance, could weaken disease surveillance and treatment programmes in countries that depend on external funding. Low educational attainment and childhood malnutrition can further compound poor health outcomes.
This makes the response a question of economic policy and international cooperation as much as healthcare.
What does the report recommend?
The commission outlines three broad priorities for governments, international institutions, industry and funders.
- Build health systems prepared for threats: Governments need stronger surveillance, prevention and emergency-response capabilities so that outbreaks and other crises can be detected and managed early.
- Invest in breakthrough innovation: Scientific research, medical technologies and other innovations can help reduce the burden of disease and improve the ability to respond to emerging dangers.
- Monitor emerging risks: Countries need better systems for identifying and tracking threats across health, climate, finance, defence and technology. Early warning and coordination can make preventive action more effective.
The report's message is not that catastrophe by 2100 is inevitable. Rather, it argues that the potential consequences are large enough to justify action well before a crisis unfolds.