Project Glasswing: What We Learned Running a Frontier Model on Our Own Codebase
hackerone.com | blog | #ai-security | #rce | #vulnerability-research | #hackerone | #frontier-models | #compositional-risk
Summary
HackerOne ran Anthropic's Mythos 5 model against its own production code; it surfaced a critical RCE spanning three safe-in-isolation commits, illustrating compositional risk that diff review misses.
- Published
- Collected
Skip to content