13:06
2026-07-10
dev.to
ai-safety
Anthropic wants to grade AI jailbreaks like CVEs. Here's the framework.
Anthropic has published a structured framework for rating the severity of AI jailbreaks, called the Cyber Jailbreak Severity (CJS) scale, which ranges from CJS-0 (Informational) to CJS-4 (Critical). T…