Hacktron AI Researchers Breach OpenAI's Security in 72 Hours!

Three researchers from Hacktron AI exploited two security weaknesses to access OpenAI employee accounts and an internal repository within 72 hours. They utilized Anthropic's Claude, reported the findings privately, and received a $6,500 bounty. OpenAI promptly patched the vulnerabilities. Notably, neither weakness involved AI flaws, and concerns about the industry's security measures are escalating. Experts argue for enhanced technical controls and regulatory clarity.