Anthropic: No Universal Jailbreak Found for AI Model

Anthropic reports that despite extensive testing, no universal method has been discovered to bypass its AI model’s safety features.