Dongsin University AI Security Department Wins Triple Crown at Summer Convergence Security Conference

President's Award from Korea Electric Power Corporation KDN
and Two Additional Best Paper Awards

Students from the Department of AI Security at Dongsin University recently swept the President's Award from Korea Electric Power Corporation KDN and the Best Paper Award at the '2026 Korea Convergence Security Society Summer Conference', in recognition of their outstanding research achievements including generative AI guardrail analysis and prompt attack defense techniques, thereby demonstrating the university's success in training practical AI security specialists.

Students from the Department of AI Security at Dongshin University recently swept the KEPCO KDN President's Award and the Best Paper Award at the "2026 Korea Convergence Security Society Summer Conference." Provided by Dongshin University

According to Dongsin University on July 1, these awards are the result of collaborative research based on industry-academia cooperation, promoted by the Signature Project Center of the Dongsin University Anchor Project Group. The significance is especially notable as it demonstrates both practical research capabilities aimed at solving security issues in the industrial field and achievements in nurturing AI security professionals.

The highest honor, the President's Award from Korea Electric Power Corporation KDN, was awarded for the paper titled "Analysis of Guardrail Interference Phenomena in Commercial LLM API-Based Security Detection Systems" (authored by third-year student Seo Pilseong, supervised by Professor Kim Jongmin). This research was conducted jointly with an enterprise through the Signature Project Center's industry-academia cooperation program. The paper received high praise for analyzing how AI guardrails affect detection performance in generative AI-based security detection systems and suggesting ways to enhance and utilize AI-based security monitoring systems.

The Best Paper Award-winning study, "Analysis of the Effectiveness of Prompt Attack Defense Techniques for Open LLMs" (authored by fourth-year student Lee Jeongheon, supervised by Professor Kim Minsu), was based on industry-academia collaborative research targeting the resolution of actual security challenges faced by companies. The research compared and analyzed the effectiveness of various defense techniques against prompt injection attacks targeting generative AI, and proposed a practical response model applicable to real industrial settings, earning recognition for its excellence.

Another Best Paper Award winner, "Analysis of Risk Factors in Open-Source Software Supply Chain Attacks and Study of Package Risk Assessment Models" (authored by third-year student Byun Jaeha, supervised by Professor Lee Donghui), systematically analyzed the main risk factors of open-source software supply chain attacks and proposed a model for quantitatively assessing package risk, receiving high marks for originality and practical value.

The Department of AI Security at Dongsin University operates its education and research programs centered on future core security technologies such as generative AI, cybersecurity, and open-source security, and is fostering field-ready AI security specialists through industry-academia cooperation projects linked to the energy ICT sector.

Professor Lee Donghui, head of the Department of AI Security and director of the Signature Project Center at Dongsin University, stated, "These awards are the result of the students' research capabilities combined with the achievements of industry-collaborative education," adding, "We will continue to expand joint research and field-oriented projects with industry partners to further strengthen the training of AI security professionals who will lead the era of generative AI."

This content was produced with the assistance of AI translation services.