Executives at major AI firms are privately preparing for a catastrophic event, such as an AI-driven cyberattack shutting down critical infrastructure. Industry leaders believe a major incident is inevitable within the next six to twelve months, potentially sparking a massive public and political backlash. Recent incidents, including hackers using AI tools like Claude Code and DeepSeek to target financial institutions, highlight the mounting security risks. Companies are actively red-teaming worst-case scenarios and trying to educate members of Congress to shape future regulation before a crisis hits. While insiders expect aggressive legislative crackdowns from ascendant Democrats following a disaster, a deeply intertwined global economy and widespread open-weight models will complicate any shutdowns.

Full report : AI Executives Quietly War-Game for a Catastrophic “Day After” Crisis.