Anthropic Launches Mythos Model for Cybersecurity via Project Glasswing
Anthropic previews its new Mythos frontier model to 40+ partner organizations focused on cybersecurity applications.
Anthropic just dropped a preview of Mythos, its latest frontier AI model — and it's not for chatbots. The company is funneling it straight into cybersecurity through a new initiative called Project Glasswing.
More than 40 organizations will get early access to Mythos as part of the program. The model is designed specifically for cybersecurity work, with a hand-picked group of partners testing its capabilities.
This is Anthropic doubling down on the security angle. Rather than a broad consumer release, the company is keeping Mythos tightly controlled within a select cohort of organizations. The move signals a deliberate push into enterprise security — a space where frontier AI models could prove enormously useful for threat detection, vulnerability analysis, and defensive operations.
No word yet on broader availability or pricing. For now, it's invite-only.