2025/11/06 08:48Google AI Introduces Consistency Training for Safer Language Models Under Sycophantic and Jailbreak Style Promptsvia MarkTechPost (author: Asif Razzaq)