LG’s proprietary artificial intelligence model, “Exaone,” is evolving into a “field-ready expert AI,” moving beyond the laboratory to deliver concrete results across various industrial sites, including new anti-hair loss materials, next-generation data center cooling materials, and financial investment analysis.

LG AI Research announced on the 8th that it is participating in ICML (International Conference on Machine Learning) 2026, the world’s most prestigious international academic conference in the machine learning and AI field, held at COEX in Seoul from the 6th to the 11th, to unveil a wide range of Exaone-based industrial innovation cases. ICML, being held in South Korea for the first time, is one of the world’s top three AI conferences and is the largest event drawing attention from the global AI academic and industrial communities.

The biggest highlight of the exhibition was the live demonstration of “Exaone Discovery,” an AI platform for discovering new materials, alongside two physical materials. One is “Rhamsydil,” a new anti-hair loss material co-developed by LG AI Research and LG Household & Health Care, and the other is an “immersion cooling fluid material” for next-generation AI data centers, co-developed with GS Caltex.

Exaone Discovery is a platform where AI autonomously reads molecular structures from literature and designs candidate substances with desired properties. LG AI Research completed patent registration for this overall R&D process earlier this year, establishing a technological barrier to entry. The application scope of this technology is expanding into cosmetics, batteries, semiconductor new materials, and new drug research.

Rhamsydil is a substance that AI identified in just one day from over 420,000 candidate materials. It has demonstrated an anti-hair loss effect without steroid-derived ingredients, with results presented at the World Congress for Hair Research, and is currently being prepared for product commercialization. The immersion cooling fluid material is a next-generation cooling solution to address the heat generation issues of power-hungry AI data centers, and the two companies plan to further expand the scope of their collaboration on discovering new materials.

In the financial sector, the finance-specialized AI agent “Exaone BI (EXAONE Business Intelligence)” drew significant attention. Exaone BI is a service that analyzes approximately 8,000 stocks listed on the South Korean and U.S. stock markets daily, providing prediction scores and analysis commentary at the level of a financial expert. Its key feature is that it is an “explainable AI,” not merely predicting stock price trends but also presenting the rationale behind the AI’s judgment.

Following the launch of a U.S. stock market prediction AI service in the global market earlier this year in collaboration with the London Stock Exchange Group (LSEG), LG AI Research recently signed a strategic business agreement with Koscom. Through this, they plan to combine Koscom’s financial data with Exaone BI’s prediction technology to build an AI prediction and analysis service specialized for the South Korean stock market.

In the data field, the AI data factory platform “Exaone Data Foundry” was unveiled. This platform is a technology that generates high-quality data with AI and automatically builds AI models specialized in professional fields. According to LG AI Research, it has achieved results in increasing data productivity by at least 1,000 times and improving quality by an average of over 20%. In a pilot project conducted with South Korea’s National Pension Service, they implemented a system that automatically builds over 10,000 pieces of specialized data per day.

LG AI Research proved its research capabilities by presenting a total of 14 papers at ICML 2026. Notably, its new material generative AI technology ranked second globally in the overall standings on “LeMat-GenBench,” a leading global performance indicator. This evaluation comprehensively verifies the stability of AI-generated crystal structures, their distinctiveness from existing materials, and the diversity of proposed candidate materials.

Since its launch in December 2020, LG AI Research has published a total of 363 papers at top-tier global AI conferences and filed a total of 838 patents, including 371 in South Korea, 243 overseas, and 224 international (PCT) applications. It is continuously strengthening its technology commercialization capabilities to apply research outcomes to actual industrial sites rather than leaving them in papers.

“Exaone is no longer a lab AI but an expert AI that finds answers in the field,” said Lim Woo-hyung, co-director of LG AI Research. “We will seize global AI leadership with a first-mover strategy for solving industrial challenges.”

Meanwhile, LG AI Research actively sought to secure global AI talent during the conference. On the 7th, it invited about 60 master’s and doctoral-level AI talents participating in ICML 2026 to an “LG AI Day” networking event, where key researchers shared the current status of LG’s AI technology development and talent nurturing plans. “AI leadership comes from the talents who create the best technology,” said Lee Hong-rak, co-director of LG AI Research. “We will leap forward as a global top-tier research hub by building a system that provides researchers with the best possible support in terms of experimental environments, data, and mentoring.”

Hyosung Accelerates AI Transformation… Establishes AI Convergence Research Institute

As LG unveiled a slew of AI commercialization achievements, Hyosung (004800) also announced on the same day that it has established an “AI Convergence Research Institute” to pursue a company-wide artificial intelligence transformation (AX) in earnest.

The Hyosung AI Convergence Research Institute plans to fully promote “applied and convergent R&D” that maximizes synergy by integrating the latest AI technologies with the business competitiveness accumulated over the past 60 years in manufacturing sectors such as heavy industry, textiles, and chemicals. The strategy is to use AI not merely as a tool for work efficiency or cost reduction, but as a core engine to lead the development of high-value-added new products and the design and execution of future new businesses for the group’s key affiliates, including Hyosung Heavy Industries (298040), Hyosung TNC (298020), and Hyosung Chemical (298000).

Maeng Sung-hyun, a professor emeritus at KAIST, has been recruited as the inaugural director (executive vice president). Director Maeng is a leading AI technology expert in South Korea, having been a professor at Syracuse University and the KAIST School of Computing, with over 30 years of research in semantic-based information retrieval, text mining, natural language processing, and language modeling. He has served as an AI advisor for Hyosung for the past eight months, consulting on the company’s digital transformation direction and conducting special lectures to enhance AI capabilities for employees.

Hyosung stated, “We will actively secure not only the core technical personnel needed for the AX transformation but also next-generation talent equipped with both industrial field expertise and AI utilization capabilities.”