BadAgent Extension

Pedro Yanes Garrido, Diego Fernandez Arias

Abstract

This paper presents an empirical study on backdoor attacks in large language model agents. We extend a recent attack framework by adding two lightweight benchmarks that measure cross-domain robustness and trigger visibility without changing the model architecture. Our approach fine-tunes AgentLM-based agents with parameter-efficient methods on operating system and web browsing tasks using multiple poisoning ratios and both visible and invisible triggers. We then evaluate the agents with

Categories

Cite This Resource

@article{llmsec202600202,
  title = {BadAgent Extension},
  author = {Pedro Yanes Garrido and Diego Fernandez Arias},
  year = {2026},
  doi = {10.4018/979-8-3373-8252-4.ch010},
  url = {https://doi.org/10.4018/979-8-3373-8252-4.ch010},
}