With Claude Opus 5, security investigators invaded OpenAI – Hacking systems
Independent security researchers used Anthropic's Claude to invade OpenAI systems, revealing gaps in the company's defenses behind ChatGPT. The news was broadcast by the Wall Street Journal. The attack was carried out by a three-member security team of the Hacktron AI startup, as part of the OpenAI Bug Bounty program. Combining two critical vulnerabilitys, [...]





