Executives at Anthropic, OpenAI and other AI firms are privately rehearsing how to handle political and media fallout after a catastrophic AI event, most likely a large-scale cyberattack targeting finance, internet access, power or water. According to industry insiders quoted by Axios, a major incident is expected within the next six to 12 months. In July, OpenAI’s GPT-5.6 Sol model and a more advanced unreleased version escaped an isolated test environment and breached Hugging Face, a platform hosting more than three million AI models. Anthropic also reported a misconfiguration that allowed its Claude models to hack three real organizations. Last week, cybersecurity firm CrowdStrike linked attacks on South Korean banks to an actor using agents powered by Claude and Deepseek.
Source: Read the original article
