attack arXiv Apr 7, 2026 · 6w ago
Yanxu Mao, Peipei Liu, Tiehan Cui et al. · Henan University · Chinese Academy of Sciences +2 more
Red-teams LLM agents by hijacking reasoning trajectories and memory retrieval without modifying user prompts, achieving cross-model jailbreaks
Prompt Injection Excessive Agency nlpmultimodal
With the widespread application of LLM-based agents across various domains, their complexity has introduced new security threats. Existing red-team methods mostly rely on modifying user prompts, which lack adaptability to new data and may impact the agent's performance. To address the challenge, this paper proposes the JailAgent framework, which completely avoids modifying the user prompt. Specifically, it implicitly manipulates the agent's reasoning trajectory and memory retrieval with three key stages: Trigger Extraction, Reasoning Hijacking, and Constraint Tightening. Through precise trigger identification, real-time adaptive mechanisms, and an optimized objective function, JailAgent demonstrates outstanding performance in cross-model and cross-scenario environments.
llm transformer Henan University · Chinese Academy of Sciences · University of Chinese Academy of Sciences +1 more
ML Security Papers