paperSeptember 2026Unreviewed

AgentLeak: Cloning Stronger LLM Agent Capabilities onto Weaker Agents Beyond Skill Stealing

Xiaoting Lyu, Yuhong Wu, Yufei Han, Shichang Liu, Liang Zhang, Bin Wang, Bin Wang, Xiaobo Ma, Wei Wang

Abstract

Large language model (LLM) agents increasingly achieve long-horizon tasks by combining foundation models with explicit skills and implicit procedural knowledge acquired through execution. The resulting task-solving capabilities have become valuable proprietary assets, raising a new security question: can a substantially weaker attacker-controlled agent acquire the capabilities of a stronger proprietary agent through limited black-box interaction? Existing skill-stealing attacks recover explicit

Cite

@misc{lyu2026agentleak,
  title = {{AgentLeak: Cloning Stronger LLM Agent Capabilities onto Weaker Agents Beyond Skill Stealing}},
  author = {Xiaoting Lyu and Yuhong Wu and Yufei Han and Shichang Liu and Liang Zhang and Bin Wang and Bin Wang and Xiaobo Ma and Wei Wang},
  year = {2026},
  month = sep,
  eprint = {2609.07131},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.07131}
}