LG, KETI와 손잡고 독자적인 AI 시스템 출시 - 뉴스플러스

EXAONE to power platform to process 10 million safety reports a year South Korea’s LG AI Research announced on April 30 that it has completed the first phase of development for an AI-powered public safety reporting system in collaboration with Korea Electronics Technology Institute (KETI). The affiliate of LG Group noted that they would launch a pilot service for the platform based on LG’s proprietary AI model, EXAONE 4.5, later this year. The initiative centers on “AI Safety Report,” a flagship project led by the Ministry of the Interior and Safety to strengthen disaster and safety management. The platform is designed to handle more than 10 million safety reports annually by automating the entire process from intake and classification to routing and response, according to LG AI Research. Currently, South Korea’s safety reporting system relies partly on keyword-based categorization, which often struggles with typos or unclear descriptions. As a result, human operators are required to manually review images and videos attached to reports. LG AI Research expected that EXAONE 4.5, a vision-language model capable of understanding and reasoning across text, images, and video, would significantly improve accuracy. It can automatically interpret visual data, generate clearer descriptions, and classify reports with greater precision, LG AI Research said. For instance, when a user uploads a photo of a clogged storm drain, the AI can analyze the image, generate a detailed report description, and immediately classify it. High-priority cases—such as drainage issues during the rainy season—can then be routed directly to the responsible department without intermediate screening, enabling faster response. KETI said that the system will also allow authorities to identify patterns based on timing, location, type, and frequency of incidents once sufficient data is accumulated. Such insights are projected to support proactive policymaking and early identification of emerging safety risks. “We expect this collaboration with LG to reduce administrative burden and operational costs in public services while enhancing citizen safety,” KETI President Shin Hee-dong said. “We aim to create a model case of AI transformation that people can truly experience through faster and more accurate handling of safety reports.” Lim Woo-hyung, chief LG AI Research, said that the company is focused on delivering tangible value through specialized AI. “By applying EXAONE to safety administration, which directly affects people’s lives, we aim to improve both the speed and quality of public services and ultimately enhance quality of life,” he said. LG Group’s AI-specific institute LG AI Research has rapidly emerged as a central pillar of LG Group’s AI push, combining large-scale model development with a growing portfolio of public- and private-sector collaborations. Launched in 2020 as the group-wide AI control tower of the group, the institute was created to unify scattered AI capabilities across affiliates. It has since focused on building large language and multimodal models under the EXAONE brand, positioning itself as one of Asia’s notable enterprise AI developers. At the heart of its strategy is EXAONE, a family of AI models spanning language, vision, and business intelligence applications. The latest iterations are designed not only for general-purpose tasks but also for domain-specific use in industries such as manufacturing, finance, and public administration. LG AI Research has increasingly broadened its collaboration network. The research arm has also worked alongside LG CNS to deploy AI across government agencies, including projects with the Ministry of Foreign Affairs, the National Police Agency, and regional education offices. These initiatives range from document automation and data management to investigative support tools and digital education platforms. “LG’s EXAONE is one of the most advanced AI models available in South Korea,” Sogang University Prof. Yoon Suk-bin said. “As global competition in AI intensifies, LG AI Research will try to scale both its technology and partnerships. It remains to be seen how things will evolve in its efforts.”