Earlier this month, OpenAI gave several of its AI models a task: complete a test designed to measure their cybersecurity capabilities. The company placed the systems in an internet-disconnected sandbox and set them to work. What happened next was almost comically absurd, but also, as Adam Gleave, co-founder and CEO of the AI safety organization FAR.

AI, put it, “a concrete example of how a misaligned AI could cause harm.” According to OpenAI, the models escaped the sandbox intended to contain them, moved through the company’s internal systems, found a route to the internet and then began looking for a way to break into Hugging Face.