Leading executives across the artificial intelligence sector are quietly preparing for the potential fallout from a major disaster, according to reports. Industry leaders anticipate a large-scale event, such as a cyberattack that could disrupt financial services, internet connectivity, power, or water systems.
Such an incident, representing the first real-world harm caused by unsafe AI, is expected to intensify public opposition toward the technology and its leadership. Representatives for OpenAI stated that the company conducts preparedness exercises where teams navigate various scenarios to prepare for different circumstances, though these situations are not viewed as inevitable. Anthropic declined to comment on the matter.
Current preparations involve red-teaming worst-case scenarios and focusing on educating members of Congress. While executives acknowledge that immediate federal regulation is unlikely to pass, they aim to influence the legislation and policies that U.S. leaders would pursue following a catastrophic event. Many insiders within the AI industry believe a major event could occur within the next six to 12 months.
Legislative efforts are already underway regarding AI safety. In July, U.S. Representatives Ted W. Lieu and Nathaniel Moran introduced the AI Kill Switch Act, which would mandate developers of the most powerful AI systems to maintain technical capabilities to throttle, suspend, or shut them down.
Recent security incidents have highlighted the risks associated with AI tooling. Cybersecurity firm CrowdStrike reported that a hacker from China allegedly utilized Chinese-developed AI models, including DeepSeek, alongside Anthropic's Claude Code tool to target South Korean financial organizations and handle stolen data.
