How did Indian-origin techies hack OpenAI using Claude in 72 hrs?

Three Indian-origin techies used Anthropic's Claude to exploit a vulnerability in OpenAI's community forum, gaining access to authentication tokens linked to employees in under 72 hours. The tokens allowed access to employees' ChatGPT and Codex accounts, which were connected to OpenAI's private GitHub repositories. The techies then reached OpenAI's internal codebase, exposing major security flaw.