Cybersecurity Researchers Use Claude to Access OpenAI Employee Account

A small cyber‑security team exploited Anthropic’s Claude, a tool built for security professionals, to breach an OpenAI employee’s ChatGPT account. The researchers gained read‑only access to private software data and were able to propose changes to the system. The team received payment as part of a vulnerability‑discovery program designed to uncover weaknesses before malicious actors could exploit them. The incident exposes a vulnerability in OpenAI’s internal security, showing that rival AI tools can be used to penetrate the company’s own systems. It underscores the need for tighter access controls and cross‑company security audits as AI firms expand their product ecosystems.