[Edaily Reporter Yun Junghoon ]LG Uplus(032640)has partnered with Seoul National University, KAIST, and Yonsei University to develop next-generation voice artificial intelligence (AI) specialized for the Korean language. The company aims to secure voice AI technology capable of understanding human speech patterns, emotions, and conversational context—as well as remembering previous content even during lengthy conversations—and to develop this into its own proprietary engine.

LG Corp. announced on the 9th that it has launched the “K-Voice AI Alliance,” an industry-academia partnership with Seoul National University, the Korea Advanced Institute of Science and Technology (KAIST), and Yonsei University.

This collaboration aims to secure core Korean voice AI technologies and reduce reliance on foreign technologies. By combining LG Corp.’s experience in operating voice AI services with the AI research capabilities of domestic universities, the alliance plans to jointly develop technologies optimized for the Korean conversational environment.

LGU+ views voice AI as a core technology for next-generation AI services and is expanding its related business. In March, the company announced its U+ Voice AI strategy during a keynote speech at MWC 2026, and since July, it has been jointly developing network-based voice AI solutions and advancing global business initiatives through a strategic partnership with global telecommunications equipment company Ericsson.

LGU+ is pooling its technological capabilities to develop its voice agent, “Ixio,” into an Agent AI and the central voice interface of the “Pigeon AI” era.

Existing voice AI systems convert a user’s speech into text, generate a response using AI, and then convert that response back into speech. Because this process involves multiple steps, it has limitations such as delayed responses and reduced naturalness in pronunciation and conversation flow. Recently, “Voice-to-Voice (V2V)” technology—which directly understands human speech and responds in speech—has been gaining attention. Since this technology can grasp not only the content of speech but also intonation, emotion, and conversational context, it enables much more natural real-time communication between humans and AI.

Within this alliance, Seoul National University is researching technology that enables conversations using context-appropriate tones and voices, as well as naturally responding while listening to the user’s speech. KAIST is responsible for developing technology that allows AI to determine whom to respond to in group conversations and to maintain contextual continuity by remembering previous content even as conversations extend. Yonsei University is applying voice AI to counseling services to analyze customer comprehension and the completeness of counseling sessions.

Based on the research findings, the participating institutions plan to establish evaluation criteria for Korean voice AI performance and jointly develop the “K-Voice Engine” and related technologies that can be applied to various services.

LGU+ intends to gradually apply the technology secured through this industry-academia collaboration to its own voice AI services and to continuously enhance the K-Voice Engine.

Lee Sang-yeop, CTO (Executive Vice President) of LGU+, stated, “Voice is a crucial means that allows anyone to use AI easily and naturally,” adding, “Through this industry-academia collaboration, we will secure K-Voice AI technology tailored to Koreans and the Korean language and provide customers with a new voice AI experience.”

He added, “We will continue to strengthen our technological competitiveness to reduce the number of people—such as the elderly—who are left behind in the AI era.”