LG AI Talk Concert 2026: ‘Expert AI’ Solving Complex Industrial Challenges
Image 1. LG AI Talk Concert 2026
Going beyond simply building good AI models to solving long-standing challenges in industrial fields—this is the path LG AI Research has walked and its core mission moving forward.
On September 14, LG AI Research hosted the ‘LG AI Talk Concert 2026’ at the LG Sciencepark Convergence Hall in Magok, Seoul. During the event, the institute presented key real-world applications of its ‘Expert AI’, showcasing technological milestones and value created across manufacturing, finance, and science since its founding.
"Expert AI that understands even the 1% exception proves its value in the field"
Image 2. Woohyung Lim, Head of LG AI Research
In his keynote speech, Woohyung Lim, Head of LG AI Research, highlighted the clear distinction between general-purpose AI and domain-specific AI, emphasizing the necessity of developing Expert AI.
"Numerous general-purpose AI models already exist and generate impressive answers. However, industrial environments demand AI that can understand complex variables and judge even the 1% exception with precision. AI must demonstrate tangible performance to prove its value in real-world operations."
Image 3. Honglak Lee, Head of LG AI Research and CSAI
Following the keynote, Honglak Lee, Head and Chief AI Scientist (CSAI) of LG AI Research, took the stage alongside George Cameron, founder of Artificial Analysis—an independent research institution that objectively analyzes and evaluates the performance, speed, intelligence, and cost of global AI models. The two discussed fair evaluation standards for global frontier AI models and shared insights into the recent rapid progress of Korean AI models and strategies to close the global technology gap. Furthermore, under the theme of the ‘Future of AI,’ they engaged in an in-depth discussion on how evaluation methodologies need to evolve and which capabilities will matter most.
Transforming Industrial Fields: Expert AI Solutions Across 3 Key Sectors
Image 4. Artificial Analysis's Co-Founder George Cameron
At the event, leaders from LG AI Research and partner organizations presented detailed accounts of tangible achievements realized through AI in manufacturing, finance, and science.
Image 5. Deep Dive Session 1 - Structured Data AI Innovation: Achieving Manufacturing AX with EXAONE Tabular
In Manufacturing, the first Deep Dive session, ‘Structured Data AI Innovation: Achieving Manufacturing AX with EXAONE Tabular’, was held in collaboration with LG Innotech. The team introduced domain-specific foundation models, including 'EXAONE Tabular', which predicts process conditions and quality, and 'EXAONE Omni-Inspect', designed to inspect products without re-training even when process conditions or visual appearances change. Through collaboration with LG Innotech, LG AI Research established a fast, accurate prediction system using minimal field data. Furthermore, by deploying a vision inspection agent that autonomously handles data sampling, labeling, and model training, the institute is accelerating the transition toward autonomous factories.
Image 6. Deep Dive Session 2 - Financial Data Innovation and Insights Driven by EXAONE Business Intelligence and Partnerships
In Finance, the second session, ‘Financial Data Innovation and Insights Driven by EXAONE Business Intelligence and Partnerships’, brought together experts from the National Pension Service (NPS), KOSCOM, and the London Stock Exchange Group (LSEG) for in-depth discussions. Key accomplishments of 'EXAONE BI', a specialized financial solution delivering end-to-end data analysis, reasoning, prediction, and explanation, took center stage. EXAONE BI is actively expanding its footprint by targeting global markets with LSEG and domestic markets with KOSCOM, while extending its coverage beyond North America, Europe, and the Middle East to encompass a wider range of asset classes.
Image 7. Deep Dive Session 3 - Beyond Automation to Autonomy: The Future of Materials R&D Presented by AI Autonomous Labs
In Science, the third session, ‘Beyond Automation to Autonomy: The Future of Materials R&D Presented by AI Autonomous Labs’, was conducted with representatives from LG Chem and GS Caltex. Together with the three experts, LG AI Research presented a vision for an autonomous laboratory ecosystem that goes beyond material design and synthesis prediction to directly perform physical experiments.
Over 100 Industrial Challenges Solved: Moving Toward Robot Foundation Models
Since its establishment, LG AI Research has solved over 100 complex industrial challenges, including battery lifespan and capacity prediction, defect detection, and supply chain optimization. These achievements are backed by top-tier research capabilities, demonstrated by 368 papers accepted at leading global AI conferences and 1,080 patents filed worldwide.
Looking ahead, LG AI Research shared its progress on developing 'Robot Foundation Models', enabling AI to perceive, decide, and act in physical environments. Prioritizing safety through proprietary core algorithms, the institute aims to advance beyond individual robot automation to create autonomous factories where an entire facility operates under a unified intelligence.
"EXpert AI for everyONE"
LG AI Research will continue its journey to ensure the true value embedded in the name EXAONE reaches both industrial fields and everyday life.