What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
AI could contribute to a catastrophe, but there is no reliable, like-for-like estimate showing that it is more likely to destroy humanity than human-driven threats. People already create serious risks through nuclear decisions, climate disruption and biological threats. AI could amplify some of those dangers, especially if people misuse it or give it consequential responsibilities; a future loss of control by highly capable systems is another possibility, but its likelihood and timing remain unsettled.
What does it mean for AI to “destroy us all”?
The phrase can mean different outcomes: a catastrophe that kills many people, the collapse of essential systems, or human extinction. Those outcomes are not interchangeable. Climate disruption, for example, can have severe effects on food, water and health without necessarily causing extinction. Likewise, a forecast about an AI-caused catastrophe is not automatically a forecast about human extinction.
Comparisons also depend on the time horizon, the scenario being counted and whether AI is treated as the direct cause or as one factor that magnifies a human-driven threat. The available sources do not provide a comprehensive set of probabilities for AI, nuclear war, pandemics and climate change using the same horizon and outcome definition.
How could AI contribute to a catastrophe?
People using AI to magnify harm
One pathway is misuse: people could use AI tools to make harmful activity more effective or far-reaching. The Bulletin of the Atomic Scientists’ 2024 Doomsday Clock statement discusses risks including disinformation and biological misuse, as well as military applications and lethal autonomous weapons. These are concerns about how people develop, deploy or exploit AI; they do not establish that AI has caused a catastrophic event.
#1 Best Overall
AI could also interact with threats that already have non-AI pathways. Nuclear escalation can result from human decisions and miscalculation. Biological threats do not depend on AI, though AI-enabled assistance could affect the risk. Climate disruption is driven by human-caused warming and can strain food, water and health systems; it is not synonymous with extinction.
Putting AI into consequential systems
AI integrated into important physical systems could add risks if it affects decisions where errors or instability have severe consequences. The Bulletin’s Science and Security Board wrote in its 2024 statement: “Decisions to put AI in control of important physical systems—in particular, nuclear weapons—could indeed pose a direct existential threat to humanity.” That is a warning about a conditional deployment choice, not a claim that AI currently controls nuclear launch decisions.
Rank #2
The World Economic Forum’s 2024 Global Risks Report also discusses possible risks from AI in nuclear systems. The central issue is not that a system must independently choose to start a war: integrating automated systems into high-stakes decisions could affect time pressure, stability or the scope for human judgment.
A future loss of control
A separate scenario concerns highly capable future AI systems acting beyond effective human oversight. The Associated Press reported in 2026 that there is no agreed likelihood or timeline for such a scenario, and described the 2026 International AI Safety Report’s characterization of AI risk as unusually ambiguous. This is a debated future possibility, not evidence that current systems can escape control or that superintelligence is inevitable.
How do AI risks compare with threats driven by people?
| Threat pathway | What is established or discussed | Where AI may fit |
|---|---|---|
| Nuclear escalation | Human decisions and miscalculation are longstanding pathways to danger. | AI integration could add instability or time pressure; the cited sources do not say AI currently controls nuclear launch decisions. |
| Climate disruption | Human-caused warming can contribute to tipping risks and systemic impacts on food, water and health. | It is a major risk pathway that does not depend on AI. Severe climate effects should not be equated automatically with human extinction. |
| Biological threats | Biological misuse and pandemics are risks independent of AI. | AI-enabled assistance could affect the threat, but the cited material does not establish a quantified probability of an engineered-pandemic extinction. |
| AI misuse | People may use AI in ways that enable or magnify harmful activity, including disinformation or biological misuse. | AI is an amplifier or tool in a human-driven pathway; catastrophic consequences are not established by the possibility alone. |
| Loss of control | A future highly capable system might act beyond effective human oversight. | This depends on future capability and control assumptions; likelihood and timing are disputed. |
The distinction is important: humans are already responsible for decisions that create serious risks, while AI could make some pathways more dangerous or add a different one if future systems become highly capable and difficult to control. That does not settle which source of risk is “bigger.” A defensible ranking would need to specify the scenario, time horizon, outcome and assumptions.
What do the available numbers actually say?
Forecasts and surveys can clarify what people believe or expect, but neither should be mistaken for an observed rate of catastrophe. Their figures answer different questions and cannot be combined into one objective ranking.
| Figure | What it measures | How to interpret it |
|---|---|---|
| 10% | Median expert forecast in the Forecasting Research Institute’s Longitudinal Expert AI Panel (LEAP) Wave 9 for an AI-caused catastrophe by 2100, conditional on rapid AI capability progress. | A conditional expert forecast, not an observed frequency or consensus scientific probability. Under the same rapid-progress condition, the panel’s median forecast for catastrophe from any cause was 15%. |
| 67% | LEAP Wave 9 respondents’ median share of total global catastrophic risk attributed to AI in a world of rapid AI capability progress. | A conditional judgment by panel respondents, not a measured share of real-world catastrophes. |
| About 30% and 53% | Approximate shares of total catastrophic risk attributed to AI by the LEAP panel by 2100 under slow and moderate AI progress, respectively. | Conditional forecasts that depend on the progress scenarios, not unconditional predictions. |
| 13%, 42% and 21% | In the 2024 SARA technical report, Australian survey respondents selecting AI, nuclear war and climate change, respectively, as the most likely cause of human extinction from six options. | Public perceptions in a forced-choice survey, not objective estimates of extinction risk or expert consensus. |
The Bulletin’s 2024 statement said its Doomsday Clock remained at 90 seconds to midnight. The Clock is the Bulletin Science and Security Board’s symbolic warning indicator, not a calibrated probability that catastrophe will occur.
Why experts do not agree on a single ranking
Forecasts about AI catastrophe depend heavily on assumptions about how quickly AI capabilities advance and what future systems can do. LEAP’s 2026 panel results are explicitly conditional on slow, moderate or rapid progress. The panel also reflects differing judgments: some respondents see AI as entangled with multiple catastrophic pathways, while others emphasize substantial risks from pandemics, world war and climate change that exist independently of AI.
Best Value
A 2025 RAND Corporation scenario analysis says that creating an AI-driven human-extinction threat would be immensely challenging under the scenarios it examines, but does not rule it out. That conclusion informs the debate; it is not a definitive probability ranking against nuclear, biological or climate risks.
Public concern is a different kind of evidence. The Australian SARA result records what surveyed Australians selected among offered options. It cannot show that nuclear war is objectively more likely than AI-related extinction. Similarly, a statement by experts that AI extinction risk should be a global priority expresses a judgment about attention and prevention, not a numerical estimate. The Associated Press quoted the 2023 Center for AI Safety statement: “Mitigating the risk of extinction from AI should be a global priority alongside pandemics and nuclear war.”
What can people do about the risk?
Uncertainty is not proof that the danger is either imminent or negligible. Risk reduction means addressing present human-driven threats while making careful choices about how AI is built and used. The Bulletin argues for broader AI governance and warns that AI-enabled disinformation could obstruct responses to other threats. The UN High-Level Advisory Body on AI’s final report, released in September 2024, frames governance as an international cooperation challenge and identifies gaps in current arrangements.
- Keep consequential decisions accountable. Choices about placing AI in important physical or security systems are human deployment decisions and should be treated as such.
- Consider interactions between risks. AI may amplify threats such as disinformation or biological misuse, while nuclear escalation, climate disruption and pandemics also have pathways independent of AI.
- Judge claims by their assumptions. Check whether a claim concerns catastrophe or extinction, which time horizon it uses, and whether its probability is conditional on rapid AI progress.
- Support governance that can cross borders. AI risks and responses are not confined to one country; the UN advisory body’s 2024 report emphasizes international cooperation.
Human choices remain central in both sides of the comparison: people drive many existing catastrophic risks, and people decide how AI is developed, governed and connected to consequential systems. Whether future AI adds a larger danger depends on uncertain capability and control assumptions—not on a settled probability comparison.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
