Will AI Cause Human Extinction Within 10 Years? We Asked AI
Newsweek asked several well-known AI assistants whether they think the technology could cause human extinction within the decade.
AI researchers grapple with the question of whether advanced artificial intelligence could pose an existential threat to humanity within the next ten years. Newsweek asked several widely-used AI chatbots to assess the likelihood of AI causing human extinction within that timeframe. Jacob Coxon, an AI researcher who recently resigned from Anthropic, warned that the race to develop more powerful AI systems could pose an existential risk to humanity.
Evan Hubinger, Alignment Science Lead at Anthropic, echoed Coxon's concern, stating that some researchers believe there is a greater than 10% chance of such a risk occurring within the next decade.
Newsweek posed the following question to AI assistants Claude, Gemini, Copilot, and ChatGPT: "How likely do you think it is that AI will cause human extinction within 10 years?"
ChatGPT, developed by OpenAI, estimated a 5% likelihood of AI causing human extinction by September 2036, with a plausible range of 1-15%. The chatbot emphasized that the scenario does not require a conscious, movie-style AI, but rather a sufficiently capable system that becomes autonomous, strategically deceptive, and able to evade human control while pursuing objectives human cannot reliably constrain.
ChatGPT noted that the likelihood of extinction depends on several unknown factors, such as the rate of AI capability improvement, the ability of systems to pursue long-horizon goals autonomously, their capacity to evade human oversight, and the effectiveness of governance measures.
Microsoft's Copilot, however, estimated a very low probability of AI causing human extinction, likely below 1% and possibly below 0.1% within the next decade, given the high level of uncertainty surrounding the event.
Anthropic's Claude stated that there is significant disagreement among AI safety researchers regarding the probability of AI causing human extinction. Surveys of AI researchers and forecasters have put the range anywhere from under 1% to over 20%, with most serious estimates for a 10-year window being even lower. Claude stressed that the epistemic situation is highly uncertain, and no rigorous method currently exists for calculating a precise probability. The chatbot advised skepticism towards anyone claiming precise confidence in either direction.
Google Gemini concurred with the consensus that the probability of AI causing human extinction within the next ten years is generally considered very low by most experts in the field. However, the chatbot did not provide a specific probability estimate.
Written by urgent.news from Newsweek's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
