Hackers breach OpenAI using Claude tools, gaining access to employee accounts

Hackers have breached OpenAI's systems using tools from Claude, gaining access to employee accounts and the company's internal codebase. The hackers initiated a harmless pull request as proof of the breach. The incident reportedly involved the use of Claude tools to gain unauthorized access to OpenAI's systems. The breach allowed the hackers to access sensitive areas of the company's internal operations. The hackers demonstrated their access by making a pull request, which was apparently harmless. The breach highlights a potential vulnerability in OpenAI's security measures.