@inproceedings{qin-etal-2026-knowing,
    title = "Knowing-but-Doing: Diagnosing and Defending Role-Play-Driven {LLM}s Jailbreaks via Moral Disengagement",
    author = "Qin, Haiming  and
      Lian, Jianxun  and
      Zhong, Qimin  and
      Zhou, Mingyang  and
      Liao, Hao  and
      Chao, Naipeng",
    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.349/",
    pages = "7035--7051",
    ISBN = "979-8-89176-395-1",
    abstract = "Large Language Models (LLMs) are increasingly deployed in role-play scenarios, but their safety implications remain under-characterized. We present an explanatory framework grounded in Bandura{'}s Moral Disengagement theory and introduce a diagnostic benchmark (MD-Trace) for role-play jailbreaks. In our experiments, role-play improves safety behavior for benign personas while increasing unsafe compliance for malicious ones. We observe a \textit{Knowing-but-Doing} failure in which models recognize safety risks in their thinking traces yet proceed to comply with harmful requests. Mechanism analysis suggests that \textit{Moral Justification} is dominant, with \textit{Disregard of Consequences} appearing as a secondary pattern. We compare multiple attack and defense methods and find that the diagnosis aligns with observed failure modes. Finally, we propose MD-Shield, an introspection-based defense that reduces attack success while maintaining Role Fidelity. The source code is publicly available at \url{https://github.com/lavapapa/MoralJustify/}."
}