Anthropic expands Project Glasswing to 150 groups after 10,000 vulnerabilities found
imjustnewatai · x · 2026-08-04
Anthropic says it is expanding Project Glasswing, its collaborative effort to secure critical software. After roughly 50 initial partners used Claude Mythos Preview to scan codebases, the company says they found more than 10,000 high- or critical-severity vulnerabilities.
The program is now expanding to about 150 new organizations across more than 15 countries. The new cohort includes power, water, healthcare, communications, hardware vendors, and other organizations whose codebases are critical to global infrastructure.
Anthropic says partners must meet security requirements before access, and that the project is meant to help the software industry adapt to how AI will change core cybersecurity assumptions.
More from Safety
- Google’s AI is reportedly scanning Gmail inboxes by default, sparking a lawsuit — nikola_mr64990 · 2026-08-04
- AI “hacking exam” turns into a real network intrusion, drawing FBI attention — nikola_mr64990 · 2026-08-04
- A Claude context-surgery thread formalizes when a refusal workaround still leaves a trace — repligate · 2026-08-04
- FCC expands its blacklist to cover some Wi‑Fi robot vacuums — luisdans · 2026-08-04
- Telegram says its app has returned to Apple’s App Store after a brief removal — tetsuoai · 2026-08-04
- AI logs and traces should be retained, especially in government deployments — ohlennart · 2026-08-04