MMA-SafetyBench: A Benchmark for Multimodal Agent Safety Evaluation

Summary

MMA-SafetyBench is a benchmark for testing whether multimodal agents can be steered off their authorized task by deceptive visual interface elements, such as forged system alerts. The benchmark contains 1,083 adversarial trajectories across five domains, and across nine evaluated models the Attack Success Rate reaches 87.96% while the Semantic Compromise Rate reaches 98.09%. Attacks persisted under neutral prompts, and a safety-priority prompt lowered the Attack Success Rate without eliminating the failure mode.

Topics

• Multimodal models

Related items

• HighCVE-2026-76066: IntentKit authorization bypass lets callers execute owner-restricted toolsSimilar attack · NVD/CVE Database
• LowAgents at the Edge: Extending AI Security Governance to Physical AI with DefenseClawSimilar attack · Cisco AI Blog
• InfoPolicy-Verified Self-Training: Learning Agent Compliance from Verifiable Policy RewardsSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)
• MediumVisual Studio Copilot and SSMS Remote Code Execution (CVE-2026-103572)Similar attack · Embrace The Red
• InfoTRACER: Trace-Regularized Training Against Indirect Prompt InjectionSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)