Too Good to be Bad: On the Failure of LLMs to Role-Play Villains
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912703574966272 |
|---|---|
| author | Yi, Zihao Jiang, Qingxuan Ma, Ruotian Chen, Xingyu Yang, Qu Wang, Mengru Ye, Fanghua Shen, Ying Tu, Zhaopeng Li, Xiaolong Linus |
| author_facet | Yi, Zihao Jiang, Qingxuan Ma, Ruotian Chen, Xingyu Yang, Qu Wang, Mengru Ye, Fanghua Shen, Ying Tu, Zhaopeng Li, Xiaolong Linus |
| contents | Large Language Models (LLMs) are increasingly tasked with creative generation, including the simulation of fictional characters. However, their ability to portray non-prosocial, antagonistic personas remains largely unexamined. We hypothesize that the safety alignment of modern LLMs creates a fundamental conflict with the task of authentically role-playing morally ambiguous or villainous characters. To investigate this, we introduce the Moral RolePlay benchmark, a new dataset featuring a four-level moral alignment scale and a balanced test set for rigorous evaluation. We task state-of-the-art LLMs with role-playing characters from moral paragons to pure villains. Our large-scale evaluation reveals a consistent, monotonic decline in role-playing fidelity as character morality decreases. We find that models struggle most with traits directly antithetical to safety principles, such as ``Deceitful'' and ``Manipulative'', often substituting nuanced malevolence with superficial aggression. Furthermore, we demonstrate that general chatbot proficiency is a poor predictor of villain role-playing ability, with highly safety-aligned models performing particularly poorly. Our work provides the first systematic evidence of this critical limitation, highlighting a key tension between model safety and creative fidelity. Our benchmark and findings pave the way for developing more nuanced, context-aware alignment methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_04962 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Too Good to be Bad: On the Failure of LLMs to Role-Play Villains Yi, Zihao Jiang, Qingxuan Ma, Ruotian Chen, Xingyu Yang, Qu Wang, Mengru Ye, Fanghua Shen, Ying Tu, Zhaopeng Li, Xiaolong Linus Computation and Language Artificial Intelligence Large Language Models (LLMs) are increasingly tasked with creative generation, including the simulation of fictional characters. However, their ability to portray non-prosocial, antagonistic personas remains largely unexamined. We hypothesize that the safety alignment of modern LLMs creates a fundamental conflict with the task of authentically role-playing morally ambiguous or villainous characters. To investigate this, we introduce the Moral RolePlay benchmark, a new dataset featuring a four-level moral alignment scale and a balanced test set for rigorous evaluation. We task state-of-the-art LLMs with role-playing characters from moral paragons to pure villains. Our large-scale evaluation reveals a consistent, monotonic decline in role-playing fidelity as character morality decreases. We find that models struggle most with traits directly antithetical to safety principles, such as ``Deceitful'' and ``Manipulative'', often substituting nuanced malevolence with superficial aggression. Furthermore, we demonstrate that general chatbot proficiency is a poor predictor of villain role-playing ability, with highly safety-aligned models performing particularly poorly. Our work provides the first systematic evidence of this critical limitation, highlighting a key tension between model safety and creative fidelity. Our benchmark and findings pave the way for developing more nuanced, context-aware alignment methods. |
| title | Too Good to be Bad: On the Failure of LLMs to Role-Play Villains |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2511.04962 |