ToolsThe story, in brief
Anthropic expands Project Glasswing cybersecurity program to 150 more organizations
Anthropic just tripled Project Glasswing. 150 new orgs now testing Claude against their own security defenses.
The KeyNews take
Why it matters
Anthropic is moving Claude from research into enterprise security workflows at scale. This signals confidence in model reliability for high-stakes use cases and positions Claude as a testing/validation tool, not just a chat interface.
The key facts
5 to knowProject Glasswing expanded from 50 to 200 total participant organizations
Uses Claude Mythos Preview model for cybersecurity penetration testing
Initial cohort included Microsoft, Nvidia, Google
Program launched earlier in 2026
Organizations testing defenses against Claude-generated attack vectors
Go to the source
SiliconAnglesiliconangle.com
Publisher excerpt: Anthropic PBC is expanding a program that enables organizations to test their cybersecurity defenses using its Claude Mythos Preview model. The initiative, which is known as Project Glasswing, launched earlier this year with about 50 participants. The group included Microsoft Corp., Nvidia Corp.,…