Indian researchers hacked OpenAI systems, earned $6,500 bounty

Three Indian-origin researchers from Hacktron AI breached OpenAI's systems in under 72 hours, accessing employee ChatGPT, Codex accounts, and a private GitHub repository. Exploiting vulnerabilities in Discourse's image-processing software, they used Claude AI to assist. Spending under $3,000, they earned a $6,500 bug bounty. OpenAI patched the issue and thanked the team.