[Edaily Reporter Yun Junghoon ]KTCorporation(030200)has partnered with researchers from Seoul National University and the Korea Advanced Institute of Science and Technology (KAIST) to secure next-generation artificial intelligence (AI) core technologies, such as autonomous agents and prompt compression.

KTCorporation announced on the 30th that it held the “Seoul National University-KTCorporation-KAIST Industry-Academia Joint Research Final Results Workshop” at the KT Umyeon Research Center in Seocho-gu, Seoul.

Approximately 50 researchers from KTCorporation, Seoul National University, and KAIST attended the workshop. Participants shared the final results of the industry-academia joint research conducted over the past year and explored ways to apply these technologies to actual business operations and services.

Through this research, KTCorporation focused on developing the following areas in collaboration with Seoul National University: △ situational context understanding and inference for autonomous agents; △ RAI (Responsible AI) evaluation criteria and reliability improvements; △ acquisition of Korean-specific datasets and development of behavioral prediction models; and △ a reinforcement learning framework for user feedback.

Additionally, in collaboration with KAIST, KT researched prompt compression and optimization technologies to reduce the input processing burden on large language models (LLMs) and laid the groundwork for significantly improving the inference efficiency of AI models and agent services.

KTCorporation plans to immediately proceed with technology validation and internalization to ensure these achievements translate into enhanced competitiveness in actual models, platforms, and services. Furthermore, the company will identify follow-up projects that reflect real-world technological needs and expand research and development (R&D) collaboration into areas such as agent-based AI, RAI, physical AI, and token efficiency.

Lee Jae-wook, Director of the Seoul National University AI Research Institute, stated, “This joint research is significant in that it connects the university’s research capabilities with real-world technical challenges in the industry and jointly validated the practical applicability of the research results.” He added, “We will continue to develop an industry-academia collaboration model through our partnership with KTCorporation, ensuring that academic achievements lead to AI technology innovation and enhanced industrial competitiveness.”

Kim Ki-eung, a professor at the Jae Chul Kim AI Graduate School at KAIST, emphasized, “The competitiveness of AI technology depends not only on model performance but also on its efficiency and applicability in real-world environments.” He added, “We expect the technology secured through this joint research to serve as a foundation for improving the processing efficiency of large-scale language models and expanding the use of AI in industrial settings.”

Park Jae-hyung, Executive Vice President and Head of the KTCorporation AX Future Technology Institute, said, “Through our collaboration with Seoul National University and KAIST, we have jointly developed the core AI technologies that KTCorporation requires.” He added, “Building on these achievements, we will expand our R&D collaboration model to grow alongside universities and continue to develop it so that the research results translate into KTCorporation’s AI competitiveness.”