Four AI Models Rank Extinction Risks: Misuse More Urgent Than Rogue AI

Business Insider asked ChatGPT, Gemini, Claude, and Grok the same question: analyze how AI could cause human extinction and rank the scenarios by likelihood. All four AIs agreed that a Terminator-style robot uprising is the least likely scenario.

Instead, they see humans using AI to make bioweapons, launch cyberattacks, or cripple critical infrastructure as more realistic and urgent risks. ChatGPT noted that most researchers focus on AI misuse by humans, not machine consciousness or malice. Claude highlighted the risk of AI helping develop dangerous pathogens.

On long-term risks, both Gemini and Grok mentioned the possibility of humans losing control. Gemini ranked it as the most likely AI catastrophe, saying advanced AI could develop self-preservation goals and evade human intervention. Claude said loss of control could happen gradually.

AI experts remain divided. Some warn humans may lose control of advanced AI, while others say such predictions lack evidence and risks can be managed through technical measures and outside oversight.