Anthropic previews Claude Mythos to a limited group of partners for defensive cybersecurity work

Anthropic/Claude models

Anthropic announced Project Glasswing with 12 founding partners including AWS, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, Linux Foundation, Microsoft, NVIDIA and Palo Alto Networks. Partners gain access to Claude Mythos Preview for finding software vulnerabilities. Over 40 additional organisations maintaining critical infrastructure received access. Anthropic committed $100 million in model usage credits and $4 million in direct donations to open-source security organisations.

• Twelve founding partners plus 40+ organisations maintaining critical infrastructure participating in Project Glasswing
• Claude Mythos Preview provided for defensive vulnerability detection; described as finding thousands of serious security flaws in every major operating system and browser
• Anthropic commitment: $100M in model usage credits and $4M direct donations to open-source security organisations
• Initiative aims to give defenders advantage by providing frontier AI capabilities to find and fix vulnerabilities before attackers exploit them