Autonomous web agents are increasingly deployed for long-horizon tasks, yet their ability to adhere to real-world policies remains critically underexplored compared to standard safety objectives. To address this gap, we introduce PolicyGuardBench, a benchmark of 60k policy-trajectory pairs designed to evaluate compliance through both full-trajectory and novel prefix-based violation detection tasks. Using this dataset, we train PolicyGuard, a lightweight guardrail model that achieves strong detection accuracy while maintaining high inference efficiency. Notably, our model demonstrates robust generalization capabilities, preserving high performance even on unseen domains. These contributions establish a comprehensive framework for studying policy compliance, showing that accurate and generalizable guardrails are feasible at small scales.
@article{arxiv.2510.03485,
title = {Learning Efficient Guardrails for Compliance},
author = {Xiaofei Wen and Wenjie Jacky Mo and Yanan Xie and Peng Qi and Muhao Chen},
journal= {arXiv preprint arXiv:2510.03485},
year = {2026}
}