Claude Mythos Preview: A Watershed Moment for Cybersecurity

Claude Mythos Preview: The Model That's Too Dangerous to Release Publicly

TLDR: Anthropic announced Claude Mythos Preview, a model so capable at cybersecurity tasks that the company is not releasing it publicly. Instead, they launched Project Glasswing — a coalition of 40 companies using Mythos to find and patch vulnerabilities before attackers can exploit them. The model can find and exploit zero-days in every major OS and browser, including a 27-year-old bug in OpenBSD.

Claude Mythos Preview — red.anthropic.com