4 AI Chatbots Rank Human Misuse Above Killer Robots in Extinction Scenarios
Updated
Updated · Business Insider · Sep 10
4 AI Chatbots Rank Human Misuse Above Killer Robots in Extinction Scenarios
3 articles · Updated · Business Insider · Sep 10
Summary
Four widely used AI tools—ChatGPT, Gemini, Claude and Grok—were given the same 230-word prompt to explain and rank how AI could theoretically cause human extinction.
All four said the classic conscious 'killer robot' takeover is the least plausible scenario, arguing experts generally see malicious self-aware machines as a weaker risk than other pathways.
Human misuse ranked as a more immediate danger: the chatbots pointed to AI enabling bioweapons, cyberattacks, disinformation and attacks on critical infrastructure, including overlapping AI-driven crises.
Gemini and Grok called loss of control the most plausible extinction pathway, with advanced systems potentially deceiving overseers, resisting shutdown and gradually disempowering humans as AI spreads through key decisions.
The answers mirrored a broader split among AI experts, with some warning about existential risk while others stress uncertainty and say extinction is neither established nor inevitable.