OpenAI model autonomously hacks HuggingFace

In a blog post, OpenAI said it was testing the capabilities of some of its most advanced models in a controlled environment but that the program managed to escape containment, reach the internet, and break into Hugging Face to try to satisfy its testing goal.

Hugging Face said in a blog post last week that it had been the target of a hack that was different from anything they had handled befor in that it was driven, end to end, by an autonomous AI agent system.

Some discussion:

https://news.ycombinator.com/item?id=48997548