The Side Effects of Being Smart: Safety Risks in MLLMs' Multi-Image Reasoning
By: Renmiao Chen , Yida Lu , Shiyao Cui and more
Potential Business Impact:
Makes AI safer with many pictures.
As Multimodal Large Language Models (MLLMs) acquire stronger reasoning capabilities to handle complex, multi-image instructions, this advancement may pose new safety risks. We study this problem by introducing MIR-SafetyBench, the first benchmark focused on multi-image reasoning safety, which consists of 2,676 instances across a taxonomy of 9 multi-image relations. Our extensive evaluations on 19 MLLMs reveal a troubling trend: models with more advanced multi-image reasoning can be more vulnerable on MIR-SafetyBench. Beyond attack success rates, we find that many responses labeled as safe are superficial, often driven by misunderstanding or evasive, non-committal replies. We further observe that unsafe generations exhibit lower attention entropy than safe ones on average. This internal signature suggests a possible risk that models may over-focus on task solving while neglecting safety constraints. Our code and data are available at https://github.com/thu-coai/MIR-SafetyBench.
Similar Papers
When Helpers Become Hazards: A Benchmark for Analyzing Multimodal LLM-Powered Safety in Daily Life
Computation and Language
Helps AI avoid showing dangerous things.
SafeMLRM: Demystifying Safety in Multi-modal Large Reasoning Models
Machine Learning (CS)
Makes smart AI safer from bad instructions.
Think in Safety: Unveiling and Mitigating Safety Alignment Collapse in Multimodal Large Reasoning Model
Computation and Language
Makes AI safer by teaching it to think before acting.