Breaking the Script: Do Role-Playing Agents Maintain Goal Alignment under Distraction?

A study presented at the 2026 Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL) finds that large language models used as role-playing agents exhibit a trade-off between conversational responsiveness and goal adherence when distracted by users. Researchers Dongxu Lu, Albert Gatt, and Johan Jeuring evaluated Gemini-2.0-Flash, Llama-3.3-70B-Instruct, and Llama-3.1-8B-Instruct, showing that larger models engage more with distractor topics while the smaller model rigidly adheres to goals. The findings highlight the need to explicitly manage this trade-off to maintain goal alignment in role-playing agents.

Abstract As large language models LLMs are increasingly deployed as role-playing agents in educational and professional training simulations, their susceptibility to user-induced distraction threatens their pedagogical utility. We formalise goal-competing distraction as a controlled evaluation paradigm and introduce a simulation framework that captures both immediate reactions and multi-turn trajectories under targeted distraction, using an LLM-based user simulator. Building upon this framework, we evaluate agent behaviour across three models: Gemini-2.0-Flash, Llama-3.3-70B-Instruct, and Llama-3.1-8B-Instruct. Our findings reveal a critical trade-off dependent on model scale. While larger models tend to remain socially responsive and more frequently engage with distractor topics, the smaller model shows rigid goal adherence by resisting and rejecting distraction. Although redirection is the most common initial response, subsequent trajectories diverge substantially. The inclusion of explicit dialogue state demonstrates model-dependent effects, acting as a stabilising anchor for smaller models but providing limited benefit for larger ones. These results suggest that maintaining goal alignment in role-playing agents requires explicitly managing the trade-off between conversational responsiveness and goal adherence.- Anthology ID: - 2026.sigdial-1.8 - Volume: Proceedings of the 27th Annual Meeting of the Special Interest Group on Discourse and Dialogue /volumes/2026.sigdial-1/ - Month: - August - Year: - 2026 - Address: - Atlanta, Georgia, USA - Editors: Jinho D. Choi /people/jinho-d-choi/ , Yun-Nung Chen /people/yun-nung-chen/ , Kotaro Funakoshi /people/kotaro-funakoshi/ , Ali Emami /people/ali-emami/ - Venue: SIGDIAL /venues/sigdial/ - SIG: SIGDIAL /sigs/sigdial/ - Publisher: - Association for Computational Linguistics - Note: - Pages: - 107–123 - Language: - URL: https://aclanthology.org/2026.sigdial-1.8/ https://aclanthology.org/2026.sigdial-1.8/ - DOI: - Cite ACL : - Dongxu Lu, Albert Gatt, and Johan Jeuring. 2026. Breaking the Script: Do Role-Playing Agents Maintain Goal Alignment under Distraction? https://aclanthology.org/2026.sigdial-1.8/ . In Proceedings of the 27th Annual Meeting of the Special Interest Group on Discourse and Dialogue , pages 107–123, Atlanta, Georgia, USA. Association for Computational Linguistics. - Cite Informal : Breaking the Script: Do Role-Playing Agents Maintain Goal Alignment under Distraction? https://aclanthology.org/2026.sigdial-1.8/ Lu et al., SIGDIAL 2026 - PDF: https://aclanthology.org/2026.sigdial-1.8.pdf https://aclanthology.org/2026.sigdial-1.8.pdf