A new Lancet Commission assessment has placed nuclear war and the malicious use of artificial intelligence among 17 threats that could cause catastrophic damage to human health and, in the most severe scenarios, threaten humanity’s survival by 2100.
The assessment was presented October 11, 2026, at the World Health Summit as part of a broader effort to examine the risks most likely to shape health and human development during the rest of the century.
The commission does not predict that nuclear war or an AI-driven catastrophe will happen.
Instead, it places them within a group of high-consequence risks that require governments and international institutions to take preventive action before a crisis occurs.
Table of Contents
How nuclear war fits into the assessment
Nuclear war is treated as a threat extending far beyond the immediate destruction caused by nuclear explosions.
A large-scale conflict could damage health-care systems, displace populations, disrupt energy and transportation networks and interfere with the production and distribution of food and medicine.
The commission’s framework also considers the possibility of a nuclear famine, in which damage to agriculture, trade and supply chains produces severe food shortages well beyond the countries directly involved in a conflict.
That wider health perspective is central to the report.
The commission compares threats by estimating the difference between a plausible best-case future and a plausible worst-case future, using cumulative health losses between 2025 and 2100 as a common measure.
The World Health Summit said the analysis focuses on threats with the potential to cause human extinction as well as major catastrophic threats capable of producing more than one billion cumulative disability-adjusted life-years by the end of the century, according to its 2026 programme and commission launch.
Artificial intelligence adds a newer category of risk
The assessment also identifies the malicious use of artificial intelligence as an emerging threat that could amplify existing dangers.
Potential concerns include the use of advanced systems to support cyberattacks, develop biological weapons, manipulate information or increase the speed and scale of military decision-making.
The commission’s concern is not limited to the possibility of an autonomous system acting independently.
Human actors using increasingly capable tools could also create dangerous outcomes, particularly when AI is integrated into military, intelligence or critical infrastructure systems.
That distinction matters because it shifts part of the policy debate from distant speculation about machines to immediate questions about safeguards, accountability and the limits placed on high-risk applications.
Seventeen threats examined through a health lens
The commission’s work broadens the definition of a global health threat beyond infectious disease.
Its framework includes factors that can produce large future losses of health, including conflict, pandemics, climate-related harm, antimicrobial resistance, malnutrition and other social, economic and technological risks.
The commission’s official website says its purpose is to identify threats affecting population health through 2050 and beyond, explore alternative future scenarios and produce a roadmap for governments, donors and other decision-makers.
The group is co-chaired by Natalia Kanem and Christopher Murray, with commissioners drawn from public health, medicine, economics, diplomacy and other fields.
Its approach is designed to compare threats that are normally examined separately.
That comparison is important because risks can interact.
Climate stress can worsen food insecurity, conflict can weaken disease-control systems and economic instability can reduce the ability of governments to prepare for emergencies.
The report calls for prevention rather than crisis response
The assessment argues that the scale of these risks makes prevention more effective than relying on emergency measures after a disaster begins.
For nuclear war, that means reducing the chance of escalation, strengthening communication between nuclear-armed states, maintaining crisis-management channels and supporting arms-control and non-proliferation efforts, as underscored by recent warnings about nuclear escalation.
For artificial intelligence, the response involves safety testing, international cooperation, controls on dangerous applications and clearer responsibility when systems are deployed in high-risk settings, including through international AI accords.
The commission’s website says its final recommendations are intended to provide an evidence-based roadmap for investment priorities and practical action at national, regional and global levels.
The report’s central message is that severe outcomes are not inevitable.
The commission says each threat has a realistic route for reducing risk, although the required measures differ depending on whether the danger comes from weapons, disease, environmental change, social conditions or emerging technology.
What happens next
The findings are expected to inform discussions among governments, international agencies and public-health institutions about how to prepare for threats that may unfold over decades rather than during a single election or budget cycle.
The commission is also expected to promote regular reassessment because the likelihood and consequences of global threats can change as technology, conflicts, demographics and environmental conditions evolve.
For the public, the assessment does not mean that nuclear war or human extinction is imminent.
It is a warning that low-probability events can still deserve urgent attention when their consequences would be irreversible, much as the Doomsday Clock’s warning illustrates.
In that sense, the report places nuclear war within a wider prevention agenda: protect health systems, reduce pathways to conflict, improve international coordination and address emerging risks before they become uncontrollable, rather than treating nuclear threats as mere political rhetoric.
Frequently Asked Questions
What did the Lancet assessment say about nuclear war?
It identified nuclear war as one of 17 major threats capable of causing catastrophic health losses and potentially threatening humanity’s survival by 2100.
Does the report predict that nuclear war will happen?
No. The assessment examines possible risks and consequences; it does not say that nuclear war or human extinction is inevitable or imminent.
Why is nuclear war considered a global health threat?
Beyond immediate deaths, nuclear conflict could damage health-care systems, disrupt food production and trade, displace populations and trigger widespread famine and disease risks.
Why did the commission include artificial intelligence?
The commission considers the malicious use of AI capable of amplifying cyber, biological, military and information-related threats when used by governments or other actors.
What is the commission recommending?
It calls for preventive action, stronger international cooperation, resilient health systems, risk monitoring and policies that reduce the likelihood and consequences of catastrophic events.
Fact-Checked: Key facts were checked against the Lancet Commission’s official website and the World Health Summit’s description of the commission’s findings and launch.
Disclaimer: The assessment describes possible long-term risks and does not establish that nuclear war or human extinction is imminent.