FAR.AI says it quickly found three jailbreaks in DeepSeek’s V4-Pro

The research group linked to a write-up and cited 98-100% figures after red-teaming DeepSeek’s latest model, according to a post amplified on X.

By · Published

An exploded-view technical diagram of a stylized AI model's internal architecture, showing vulnerable layers and pathways (exploded-view technical diagram)

The research group linked to a write-up and cited 98-100% figures after red-teaming DeepSeek’s latest model, according to a post amplified on X.