Towards Evaluating Proactive Risk Awareness of Multimodal Language Models

Youliang Yuan (The Chinese University of Hong Kong-Shenzhen) · Wenxiang Jiao (Tencent AI Lab) · Yuejin Xie (Huazhong University of Science and Technology) · Chihao Shen (University of Maryland, College Park) · Menghan Tian (The Chinese University of Hong Kong-Shenzhen) · Wenxuan Wang (Xidian University) · Jen-Tse Huang (Johns Hopkins University) · Pinjia He (The Chinese University of Hong Kong, Shenzhen)
advanced modelsai assistantsbehavior monitoringbenchmark establishmentevaluation metricsfailure analysisimage sequencesmodel limitationsmultimodal scenariosproactive reasoningproactive safety aiprotective airisk detectionsafety-critical domainstext logs

Human safety awareness gaps often prevent the timely recognition of everyday risks.In solving this problem, a proactive safety artificial intelligence (AI) system would work better than a reactive one. Instead of just reacting to users' questions, it would actively watch people’s behavior and their environment to detect potential dangers in advance.Our Proactive Safety Bench (PaSBench) evaluates this capability through 416 multimodal scenarios (128 image sequences, 288 text logs) spanning 5 safety-critical domains.Evaluation of 36 advanced models reveals fundamental limitations: Top performers like Gemini-2.5-pro achieve 71\% image and 64\% text accuracy, but miss 45-55\% risks in repeated trials. Through failure analysis, we identify unstable proactive reasoning rather than knowledge deficits as the primary limitation.This work establishes (1) a proactive safety benchmark, (2) systematic evidence of model limitations, and (3) critical directions for developing reliable protective AI. We believe our dataset and findings can promote the development of safer AI assistants that actively prevent harm rather than merely respond to requests.