Anthropic

Next-generation Constitutional Classifiers: More efficient protection against universal jailbreaks

Jan 9, 2026
Next-generation Constitutional Classifiers: More efficient protection against universal jailbreaks
SummaryAI, from the source post

Introduced Constitutional Classifiers++ (CC++), an evolution of the Constitutional Classifiers system for defending against LLM jailbreaks. CC++ features a two-stage cascade architecture: a lightweight probe classifier screens internal model activations for suspicious patterns, and if flagged, escalates to a more powerful probe-classifier ensemble that monitors both inputs and outputs. This system achieves a 0.05% refusal rate on harmless queries and a detection rate of 0.005 per thousand queries, with no universal jailbreaks discovered during extensive red-teaming. The system adds approximately 1% compute overhead.

More from the Modular Knowledge Compartmentalization for AI Models trail