OpenAI hacked by small team of white hat security researchers

A small team of white hat security researchers successfully hacked OpenAI. The researchers used Claude, an AI model developed by Anthropic, to achieve this. The details of the hack are discussed in an article on VentureBeat and a corresponding comments thread on Hacker News. The hack was apparently done to test the security of OpenAI's systems. The researchers' methods and findings have been shared online for discussion.