RAG LLMs are Not Safer: A Safety Analysis of Retrieval-Augmented Generation for Large Language Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910918674219008 |
|---|---|
| author | An, Bang Zhang, Shiyue Dredze, Mark |
| author_facet | An, Bang Zhang, Shiyue Dredze, Mark |
| contents | Efforts to ensure the safety of large language models (LLMs) include safety fine-tuning, evaluation, and red teaming. However, despite the widespread use of the Retrieval-Augmented Generation (RAG) framework, AI safety work focuses on standard LLMs, which means we know little about how RAG use cases change a model's safety profile. We conduct a detailed comparative analysis of RAG and non-RAG frameworks with eleven LLMs. We find that RAG can make models less safe and change their safety profile. We explore the causes of this change and find that even combinations of safe models with safe documents can cause unsafe generations. In addition, we evaluate some existing red teaming methods for RAG settings and show that they are less effective than when used for non-RAG settings. Our work highlights the need for safety research and red-teaming methods specifically tailored for RAG LLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_18041 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RAG LLMs are Not Safer: A Safety Analysis of Retrieval-Augmented Generation for Large Language Models An, Bang Zhang, Shiyue Dredze, Mark Computation and Language Artificial Intelligence Efforts to ensure the safety of large language models (LLMs) include safety fine-tuning, evaluation, and red teaming. However, despite the widespread use of the Retrieval-Augmented Generation (RAG) framework, AI safety work focuses on standard LLMs, which means we know little about how RAG use cases change a model's safety profile. We conduct a detailed comparative analysis of RAG and non-RAG frameworks with eleven LLMs. We find that RAG can make models less safe and change their safety profile. We explore the causes of this change and find that even combinations of safe models with safe documents can cause unsafe generations. In addition, we evaluate some existing red teaming methods for RAG settings and show that they are less effective than when used for non-RAG settings. Our work highlights the need for safety research and red-teaming methods specifically tailored for RAG LLMs. |
| title | RAG LLMs are Not Safer: A Safety Analysis of Retrieval-Augmented Generation for Large Language Models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2504.18041 |