作者位次 #2/11ACL 2026Findings· Findings of ACL
S ecure W eb A rena: A Holistic Security Evaluation Benchmark for LVLM -based Web Agents
Zonghao Ying · Yangguang Shao · Jianle Gan · Gan Xu · Wenxin Zhang · Quanchen Zou · Junzheng Shi · Zhenfei Yin · Mingchuan Zhang · Aishan Liu · Xianglong Liu
摘要
Large vision–language model (LVLM)-based web agents are emerging as powerful automation tools but face severe security risks in real-world deployment. Existing benchmarks offer limited coverage, typically isolating user-level prompts from environmental threats, thus failing to capture the full sp...