(SK telecom)
South Korean technology and telecom companies are accelerating efforts to capture the fast-growing voice artificial intelligence market as advances in multimodal AI sharpen competition over Korean-language services tailored to local users.
The push comes as AI models capable of simultaneously processing text, voice and images become more sophisticated, giving domestic companies an opportunity to compete against global rivals with Korean-focused language understanding.
SK telecom has emerged as one of the most aggressive players. The company recently integrated its AI agent “A.Auto” into Renault Korea’s new vehicle model Palant, powered by its Korean-language large language model A.X 4.0.
The service goes beyond basic voice commands such as navigation or music playback by accurately recognizing Korean slang and colloquial expressions, including requests to activate heated seats.
KT Corp. is targeting home entertainment with its Genie TV AI Agent service, which allows users to ask questions ranging from weather and news to science and educational topics through conversational voice interactions.
Meanwhile, LG Uplus has introduced its AI call agent “ixi-O,” designed to analyze conversational context, emotional tone and speech patterns while proactively responding to threats in real time.
Platform giants Naver and Kakao are also preparing broader voice AI expansions.
Naver is focusing on the business-to-business market through upgrades to its Naver Works Clova Note service, including automatic speaker recognition technology capable of distinguishing among multiple participants during meetings. The company plans to further enhance real-time speech recognition and summarization functions later this year as it seeks to strengthen the platform’s position as an essential workplace productivity tool.
Kakao plans to expand commercial applications of “Kanana Omni,” its integrated multimodal language model capable of understanding text, voice and images simultaneously, within the first half of this year.
The company believes its stronger Korean-language comprehension compared with some global AI models could provide a competitive edge in the domestic market.
According to Kakao, Kanana Omni currently ranks among the top Korean multimodal AI models in benchmark evaluations. The company is also conducting beta tests of its application programming interface service for a limited number of users.
Kakao recently added voice capabilities to the government-backed AI public assistant pilot service as part of broader efforts to expand its AI ecosystem.
댓글
0