On June 16, KT released the benchmark 'KSAFE-MM,' which evaluates how safely multimodal large language models (MLLMs) provide answers that reflect Korean social issues and cultural context. Co-developed with Korea University, the benchmark consists of 'KSAFE-MM-G,' which translates global common risks into the Korean cultural context, and 'KSAFE-MM-C,' which reflects Korea-specific social issues such as rental fraud and the Dokdo dispute. Based on a total of 14,135 evaluation samples—the largest number in Korea—KT verified 12 global MLLMs, including Google Gemma and Naver HyperCLOVA X. A key feature is the introduction of an automated, universal pipeline covering the entire process from collecting sensitive topics to generating jailbreak queries. The joint research team demonstrated the pipeline’s global applicability through a pilot experiment (JSAFE-MM-C) adapted to Japanese. KT expects the results can be used for safety verification, red team testing, and guardrail model evaluation. The research findings and benchmark are available on Archive and Hugging Face.