FENCE: A Financial and Multimodal Jailbreak Detection Dataset

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Kim, Mirae, Jeong, Seonghun, Kwak, Youngjun
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917284088381440
author Kim, Mirae
Jeong, Seonghun
Kwak, Youngjun
author_facet Kim, Mirae
Jeong, Seonghun
Kwak, Youngjun
contents Jailbreaking poses a significant risk to the deployment of Large Language Models (LLMs) and Vision Language Models (VLMs). VLMs are particularly vulnerable because they process both text and images, creating broader attack surfaces. However, available resources for jailbreak detection are scarce, particularly in finance. To address this gap, we present FENCE, a bilingual (Korean-English) multimodal dataset for training and evaluating jailbreak detectors in financial applications. FENCE emphasizes domain realism through finance-relevant queries paired with image-grounded threats. Experiments with commercial and open-source VLMs reveal consistent vulnerabilities, with GPT-4o showing measurable attack success rates and open-source models displaying greater exposure. A baseline detector trained on FENCE achieves 99 percent in-distribution accuracy and maintains strong performance on external benchmarks, underscoring the dataset's robustness for training reliable detection models. FENCE provides a focused resource for advancing multimodal jailbreak detection in finance and for supporting safer, more reliable AI systems in sensitive domains. Warning: This paper includes example data that may be offensive.
format Preprint
id arxiv_https___arxiv_org_abs_2602_18154
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle FENCE: A Financial and Multimodal Jailbreak Detection Dataset
Kim, Mirae
Jeong, Seonghun
Kwak, Youngjun
Computation and Language
Artificial Intelligence
Databases
Jailbreaking poses a significant risk to the deployment of Large Language Models (LLMs) and Vision Language Models (VLMs). VLMs are particularly vulnerable because they process both text and images, creating broader attack surfaces. However, available resources for jailbreak detection are scarce, particularly in finance. To address this gap, we present FENCE, a bilingual (Korean-English) multimodal dataset for training and evaluating jailbreak detectors in financial applications. FENCE emphasizes domain realism through finance-relevant queries paired with image-grounded threats. Experiments with commercial and open-source VLMs reveal consistent vulnerabilities, with GPT-4o showing measurable attack success rates and open-source models displaying greater exposure. A baseline detector trained on FENCE achieves 99 percent in-distribution accuracy and maintains strong performance on external benchmarks, underscoring the dataset's robustness for training reliable detection models. FENCE provides a focused resource for advancing multimodal jailbreak detection in finance and for supporting safer, more reliable AI systems in sensitive domains. Warning: This paper includes example data that may be offensive.
title FENCE: A Financial and Multimodal Jailbreak Detection Dataset
topic Computation and Language
Artificial Intelligence
Databases
url https://arxiv.org/abs/2602.18154