12/10/2025
•
EN
The Normalization of Deviance in AI
Explores the 'Normalization of Deviance' concept in AI safety, warning against complacency with LLM vulnerabilities like prompt injection.