Anthropic, OpenAI prepare for AI disaster that could shut down internet, power, water
Anthropic, OpenAI and other AI companies are privately war-gaming the political fallout of a catastrophic AI event, Axios reports. Executives fear a major cyberattack could knock out banking, internet, power and water, with insiders expecting an incident within six to 12 months. The firms are red-teaming worst-case scenarios and briefing Congress to shape future AI rules, with Democrats eyeing a crackdown after the November midterms.



