Anthropic warns Claude Mythos Preview finds bugs faster than developers can patch them

Anthropic says its Claude Mythos Preview model, tested with around 50 partners, found more than 10,000 high- or critical-severity vulnerabilities.

The findings reportedly helped accelerate patching at major companies and uncovered thousands of issues in open-source projects.

Only a small share of those open-source bugs has been patched so far, underscoring a growing disclosure and remediation backlog.

The report suggests advanced AI is already outpacing human security teams at finding exploitable flaws, with benefits for defense and risks for overload.