Probability of Any Global Catastrophe
Probability of Any Global Catastrophe
Experts put the chance of any global catastrophe before 2100 at 10% — and attribute half of that to AI alone.
Hover to highlight a scenario, click to show or hide itTap a scenario to show or hide it
Our Analysis
How do we interpret these results
All participant groups forecast that total catastrophic risk rises with both the time horizon and the assumed pace of AI progress. Unconditionally, the median expert assigns a probability of 1% by 2030, 4.2% by 2050, and 10% by 2100. Under Slow Progress, the corresponding expert medians are 0.75%, 3%, and 8%; under Rapid Progress, they rise to 2%, 7%, and 15%. The public is generally more pessimistic, especially under scenario-conditioned forecasts: by 2100, the median public respondent assigns probabilities of 10% under Slow Progress, 12% under Moderate Progress, and 20% under Rapid Progress. Superforecasters are slightly more optimistic than both groups, assigning unconditional medians of 0.6%, 3%, and 9%. Differences between experts and the public are statistically significant across all scenario-conditioned forecasts, though the groups converge in the unconditional 2100 forecast, where both assign a median of 10%. Dispersion remains substantial: under Rapid Progress by 2100, the expert interquartile range is 7%-25%.
Forecasts by scenario
Median forecasts from experts for each scenario
| scenario | 2030 | 2050 | 2100 |
|---|---|---|---|
| Unconditional | 1% | 4.2% | 10% |
| Slow Progress | 0.75% | 3% | 8% |
| Moderate Progress | 1% | 5% | 10% |
| Rapid Progress | 2% | 7% | 15% |
Click any cell to compareTap any cell to compare
* Denotes non-overlapping bootstrapped confidence intervals between group aggregates.
Rationale analysis
The main lines of reasoning driving the spread in estimates
Modest increase
These respondents tend to reason that AI’s footprint is becoming so broad that a genuinely AI-free catastrophe is unlikely: “AI will be heavily interwoven in all aspects of life such that almost any major decision/activity would be limited by the ‘if not for’ test.”; “AI could become instrumental in everything that happens henceforth, and there would then be no meaningful difference between Q3 and Q4.”; “The bulk of my probability mass falls on AI-caused disasters, so I am only making small adjustments here…there just aren't that many risk factors that don’t at least indirectly flow through AI.”; “Little additional probability that wouldn't be causally linked to AI, as AI takes an increasing share of catastrophic risk-relevant decision-making (pandemic prevention, asteroid detection, nuclear command & control…)” Others argue that AI will help mitigate the risk of a non-AI catastrophe occurring: “While the risk of global catastrophe are slight in almost any scenario, AI I think can actually make us safer with time, for example, by monitoring threats from out space, or by helping analyze new viruses and craft vaccines.”
Major increase
Many of these respondents reject the premise that AI will be entangled with nearly everything and treat the dominant non-AI pathways to catastrophe as carrying large standalone risks: “Pandemic risks and world war top the risks, followed by risks from climate change causing widespread crop failures or oceanic biosphere collapse. I consider AI to be a small component of overall threats to the world.”; “I consider catastrophes resulting from climate change to be the most likely, followed by wars between highly populated nuclear-armed nations.”; “Unlike the AI question, baseline risk here is already non-negligible without AI—pandemics, nuclear war, [and] climate collapse are all live on a 75-year horizon.”
Rationale examples
High-quality rationales from forecasters across the range
Modest-increase respondents
I am generally optimistic that many disasters can be either prevented or substantially mitigated through technological progress. For example, we can already predict typhoons with increasing accuracy, and in principle, we may eventually develop the ability to manage aspects of the Earth’s climate at a large scale. Advances in these capabilities will likely be strongly supported by AI. AI systems can integrate vast amounts of environmental data as contextual inputs, improving both disaster prediction and the design of mitigation strategies to reduce harm. Note that I am referring primarily to catastrophes arising from natural disasters.
The one major existential threat besides AI enabled technology is climate change. So, [this question’s wording] increases the odds slightly that a billion people may die due to increased effects of climate change. But as AI progresses in the rapid case, we may also see solutions, that prevent such catastrophes.
Major-increase respondents
Climate change appears inevitable at this point, the consequences of which could be dire for many populations. Climate change will force unwanted migration and conflict over territory as it has for millenia. Highly mobile populations have demonstrated the speed of global pandemics as a threat to dense populations. The combination of mass migration and a pandemic would certainly threaten a large percentage of the world's population in the time periods of this question.
My main model is not that we intentionally walk into extinction with a bugle and a laminated checklist. I think we are usually too self-interested and too afraid to deliberately cross the brightest red lines. The bigger risk is that we are optimistic, status-seeking, stick-poking explorers who normalize dangerous systems until the danger stops feeling dangerous. We get comfortable around reactors, pathogens, nuclear command systems, ecological fragility, AI-enabled cyber systems, and brittle global supply chains because yesterday did not end the world, and humans are very good at mistaking ‘not dead yet’ for ‘safely understood’.
I don't think contemporary climate modeling is able to internalize what could happen if certain planetary tipping points are crossed. I am not a hothouse earth believer, but I do think that we know far too little about what the actual potential climate impacts could be… Adding on the social and political instability that is likely to increase as they accumulate, I think a catastrophic population loss is more likely than not in the rapid progress scenario by 2100 because I believe that global instability will lead to a greater potential for misuse of powerful technology. In prior surveys, I have commented that rapid progress in AI would likely augment our capacity to mitigate climate change's GDP impacts, and I think this is true. But I think [mitigation] will be heavily concentrated in rich nations and will persist despite catastrophic outcomes… Postcolonial poor nations…will experience the worst outcomes while first movers on the technology will remain somewhat insulated from compounding systemic risks.