Towards Evaluating Proactive Risk Awareness of Multimodal Language Models

TL;DR

This paper introduces PaSBench, a benchmark for evaluating the proactive safety capabilities of multimodal language models across various safety-critical scenarios, revealing significant limitations in current models' ability to predict risks.

Contribution

It establishes the first proactive safety benchmark for multimodal models, providing systematic evaluation and analysis of their limitations in risk detection.

Findings

Top models achieve around 70% accuracy but miss 45-55% of risks

Failure analysis shows issues in proactive reasoning, not knowledge deficits

Dataset and benchmark promote development of safer, proactive AI systems

Abstract

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. Our dataset can be found at https://huggingface.co/datasets/Youliang/PaSBench.