30 days with Claude Mythos Preview: How Tenable adapted our security program, and why yours is next

Tenable spent 30 days running frontier AI models against our own code. It didn’t just find bugs — it proved they’re real, with reproducible exploits.