paperSeptember 2026Unreviewed

Structurally Close, Temporally Distant: Measuring Security Exposure in Long-Horizon LLM Agents

Md Jafrin Hossain, Nur Al Hasan Haldar

Abstract

Long-horizon LLM agents interact with untrusted content, persistent memory, external state, and sensitive tools. Existing analyses often characterize attacks by the number of execution steps between malicious input and a downstream action. We show that temporal remoteness can overstate security separation in stateful agents. We introduce a provenance-aware execution graph linking agent events through deterministic state, identifier, and tool provenance, and define \emph{influence distance} $\DI$

Cite

@misc{hossain2026structurally,
  title = {{Structurally Close, Temporally Distant: Measuring Security Exposure in Long-Horizon LLM Agents}},
  author = {Md Jafrin Hossain and Nur Al Hasan Haldar},
  year = {2026},
  month = sep,
  eprint = {2609.05911},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.05911}
}