Anthropic Expands Project Glasswing to 150 Organizations in 15+ Countries, Scaling Claude Mythos
Updated
Updated · TechCrunch · Jun 2
Anthropic Expands Project Glasswing to 150 Organizations in 15+ Countries, Scaling Claude Mythos
3 articles · Updated · TechCrunch · Jun 2
About 150 new organizations across more than 15 countries gained access Tuesday to Anthropic’s Project Glasswing, extending Claude Mythos vulnerability scanning into power, water, healthcare, communications and hardware.
Claude Mythos sits at the center of the program; Anthropic says the model can uncover thousands of zero-day flaws over several weeks and is being deployed where a successful attack could be catastrophic.
Many of the new participants maintain codebases used by governments and other organizations, and Anthropic estimates a major breach at most partners could affect more than 100 million people.
Financial Times reported the expanded group includes Okta, Samsung, SK Hynix, SK Telecom, NATO and EU cybersecurity agency ENISA, spanning U.S.-aligned countries including Japan, Germany, India and South Korea.
The rollout comes a day after Anthropic confidentially filed for an IPO and as it races to build safeguards before rivals match Mythos’s capabilities, with OpenAI already testing GPT-5.5-Cyber.
As AI finds flaws faster than humans can fix them, is this defensive tool creating a more dangerous cyber landscape?
If AI can autonomously exploit decades-old bugs, can any software ever be considered truly secure again?
Project Glasswing’s AI Uncovers 6,202 Critical Vulnerabilities: How Anthropic’s Mythos Preview Is Transforming Global Cybersecurity—and Creating a Patch Crisis
Overview
Project Glasswing has rapidly expanded as of June 2026, now including around 150 organizations across more than 15 countries. This growth brings together key players from critical infrastructure sectors such as power, water, healthcare, communications, and hardware. Major companies like Verizon have joined the alliance, participating in AI safety trials to address cybersecurity risks and protect their networks. The initiative highlights the urgent need for advanced defenses as attackers increasingly use frontier AI to accelerate cyberattacks. Industry leaders like IBM are integrating new AI-era protections, showing how collaboration is essential to secure vital systems in a changing threat landscape.