AI RESEARCH
AutoRAN: Automated Hijacking of Safety Reasoning in Large Reasoning Models
arXiv CS.LG
•
ArXi:2505.10846v3 Announce Type: replace This paper presents AutoRAN, the first framework to automate the hijacking of internal safety reasoning in large reasoning models (LRMs). At its core, AutoRAN pioneers an execution simulation paradigm that leverages a weaker but less-aligned model to simulate execution reasoning for initial hijacking attempts and iteratively refine attacks by exploiting reasoning patterns leaked through the target LRM's refusals. This approach steers the target model to bypass its own safety guardrails and elaborate on harmful instructions.