Balancing Security and Performance in LLM Agents: Spotlight-Guard, a Layered Defense Against Indirect Prompt Injection

Abstract

Large Language Model (LLM)-based agents automate complex tasks by integrating external tools such as web browsers, e-mail clients, file readers, and APIs, but this same integration exposes them to indirect prompt injection (IPI) attacks, in which malicious instructions hidden in tool content hijack the agent. A central but often overlooked question is how defending against such attacks affects the LLM and its own task performance and computational efficiency. In this study, we design a comprehen

Categories

Framework mappings

Suggested from the entry's categories.

@article{demirol2026balancing,
  title = {{Balancing Security and Performance in LLM Agents: Spotlight-Guard, a Layered Defense Against Indirect Prompt Injection}},
  author = {Doygun Demirol and Murat Aydoğan},
  year = {2026},
  month = aug,
  journal = {Applied Sciences},
  doi = {10.3390/app16157662},
  url = {https://www.semanticscholar.org/paper/bdb097dcf666e8ed89f66d2c7319d47f524fc614}
}