@inproceedings{huang-etal-2026-guardian,
    title = "Guardian-as-an-Advisor: Advancing Next-Generation Guardian Models for Trustworthy {LLM}s",
    author = "Huang, Yue  and
      Zhuang, Haomin  and
      Ye, Jiayi  and
      Bao, Han  and
      Wang, Yanbo  and
      Hua, Hang  and
      Wu, Siyuan  and
      Chen, Pin-Yu  and
      Zhang, Xiangliang",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.findings-acl.292/",
    pages = "5878--5900",
    ISBN = "979-8-89176-395-1",
    abstract = "Hard-gated safety checkers often over-refuse and misalign with a vendor{'}s model spec; prevailing taxonomies also neglect robustness and honesty, yielding safer-on-paper yet less useful systems. This work introduces Guardian-as-an-Advisor (GaaA), a soft-gating pipeline where a guardian predicts a binary risk label plus a concise explanation and prepends this advice to the original query for re-inference, keeping the base model operating under its original spec. To support training and evaluation, GuardSet is constructed{---}a 208k+ multi-domain dataset unifying harmful and harmless cases with targeted robustness and honesty slices. GuardAdvisor is trained via SFT followed by RL to enforce label{--}explanation consistency. GuardAdvisor attains competitive detection accuracy while enabling the advisory workflow; when used to augment inputs, responses improve over unaugmented prompts. A latency study shows advisor inference uses below 5{\%} of base-model compute and adds only 2{--}10{\%} end-to-end overhead under realistic harmful-input rates. Overall, GaaA steers models to comply with the model spec, maintaining safety while reducing over-refusal."
}