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 extinction scenarios are hypothetical pathways in which AI contributes to the extinction of humanity. They are one part of the wider discussion of AI existential risk; related outcomes such as societal collapse or permanent loss of human agency are serious, but they are not the same as human extinction.
What counts as an AI extinction scenario?
An AI extinction scenario describes a causal pathway in which AI plays a role in humanity’s extinction. The phrase identifies a possible endpoint, not a forecast that it will happen.
The UK Government Office for Science’s AI 2030 scenarios report notes that definitions of existential risk vary. For its analysis, it uses a working definition that includes human extinction or societal collapse, while stating that the report is not government policy. For clarity, it is useful to reserve “extinction scenario” for human extinction and use “existential catastrophe” as a broader term when discussing collapse or lasting loss of human agency.
How could AI contribute to human extinction?
Two broad scenario families help distinguish the main causal pathways. In one, an AI system operates beyond human control; in the other, people use AI to help cause catastrophic harm. These are hypothetical families, not evidence that current systems can bring about extinction.
#1 Best Overall
Loss of control
A loss-of-control scenario involves one or more general-purpose AI systems operating outside anyone’s control, with no clear path to regaining it. The International AI Safety Report 2026 describes possible outcomes of such scenarios as varying in severity. Some researchers consider human marginalisation or extinction possible, but those are hypothesized outcomes, not inevitable consequences.
Expert views on how likely loss of control is differ substantially. Disagreement includes questions about future capabilities, how systems will behave, and how they will be deployed. The report does not establish a settled probability or timeline.
Rank #2
Deliberate misuse
In a misuse scenario, people use AI to enable catastrophic harm. A RAND analysis examines potential pathways involving nuclear weapons, pathogens, and geoengineering. It describes creating an extinction threat through these technologies with AI as immensely challenging, while not ruling out the possibility. These are studied pathways, not demonstrated current capabilities.
Recommended Free Tools
The two families differ in who or what drives the harm: loss of control centers on systems acting beyond human control, while misuse centers on people using AI as an aid. In either case, a claim about extinction requires a causal argument connecting the AI’s role to that endpoint; a harmful or disruptive AI application alone does not establish such a pathway.
How should these scenarios be evaluated?
A useful assessment separates the links in a proposed scenario rather than treating “AI risk” as a single, measurable outcome. Relevant questions include:
- Agent: Is the scenario driven by an AI system acting beyond control, or by people using AI?
- Enabling technology: What system or harmful technology is involved?
- Causal steps: What would have to happen for AI involvement to lead to the proposed outcome?
- Evidence: What is established about relevant present-day capabilities, and what depends on assumptions about future systems?
- Endpoint: Does the pathway end in human extinction, societal collapse, or permanent loss of human agency?
- Uncertainty: How uncertain is each step, rather than just the overall scenario?
The 2026 international report records expert disagreement about loss-of-control risks. The RAND analysis examines specific misuse pathways. These sources support comparing scenario mechanisms and evidence, but do not provide a common quantitative scale for scoring them.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What is known about likelihood and timing?
The sources cited here do not establish one authoritative probability or arrival date for AI-driven human extinction. A plausible pathway is not proof that its outcome is likely, and uncertainty should not be mistaken for certainty in either direction.
The International AI Safety Report 2026 describes substantial disagreement among experts about loss-of-control likelihood. The 2024 interim International Scientific Report says expert opinion can offer guidance where research is limited, but cannot replace research. Together, these reports are reasons to treat specific forecasts cautiously rather than repeat an unsupported probability or timeline.
Best Value
Why do people call AI extinction risk a priority?
The Center for AI Safety published the statement: “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.” This is an advocacy statement about priorities, not a measured estimate of how likely extinction is.
That distinction matters: arguing that a risk deserves attention does not establish its probability. Likewise, discussing an extinction scenario does not imply that societal collapse, disempowerment, or other severe outcomes are synonymous with extinction.
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.
