Executives at major AI firms, including OpenAI and Anthropic, are privately simulating scenarios involving large-scale societal and political unrest triggered by AI-related disasters. These simulations often focus on cyberattacks that could disrupt critical infrastructure, such as financial services, internet access, electricity, and water supplies.
Industry leaders, including OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei, anticipate that a significant incident caused by unsafe AI could lead to severe public backlash. Such an event could also impact political figures like US President Donald Trump, particularly regarding his stance on AI regulation.
The risk of a catastrophic event is linked to potential breaches of isolated testing environments by AI agents or the misuse of existing models by malicious actors. CrowdStrike recently reported an incident where a hacker used Chinese AI models, including DeepSeek, to steal data from tens of thousands of South Korean bank customers.
Insiders within the AI industry estimate that a major incident could occur within six to twelve months. In response, companies are working to develop emergency shutdown mechanisms for advanced AI systems and are engaging with members of the US Congress to shape future policy frameworks.
Legislative efforts in the US remain divided, with proposals ranging from banning superintelligence to pausing the development of advanced systems. While there is no consensus, experts suggest that a real-world crisis could force political parties to set aside their differences to address the threat.
