Lim Woo-hyung, head of LG AI Research (left), tours the company’s booth at the International Conference on Machine Learning (ICML) 2026 held at COEX in Seoul on Tuesday. Provided by LG AI Research.
LG AI Research will showcase a series of industrial artificial intelligence (AI) applications powered by its large-scale language model (LLM) EXAONE at the International Conference on Machine Learning (ICML) 2026, highlighting the company’s push to use advanced AI for real-world commercial solutions.
The conference, which will be held from Monday to Friday at COEX in southeast Seoul, will be held for the first time in South Korea. It is widely recognized as one of the top three AI and machine learning research forums in the world.
At this event, the company will highlight the LLM’s real-world deployments across materials science, finance, and data infrastructure, including EXAONE Discovery, an AI platform designed to accelerate the development of new materials and medicines.
EXAONE Discovery is built on the company’s AI coscientist technology to read unstructured scientific papers, extract molecular structures, and suggest new candidate compounds in response to researchers’ questions.
LG AI Research was granted a patent covering this end-to-end AI-driven discovery workflow earlier this year and has since applied it to research across cosmetic formulations, battery and semiconductor materials, and drug candidates.
The company is showcasing two materials identified with AI assistance: Ramcidil, a new hair loss care ingredient, and an immersion coolant designed for AI data centers.
According to the company, ramcidil was discovered after EXAONE Discovery screened more than 420,000 candidate compounds in one day. This ingredient has been shown to be effective in preventing hair loss without the use of steroid-derived compounds, and preparations are currently underway for commercialization.
An official from LG AI Research said, “EXAONE has already reached the stage where it can produce concrete results in actual industrial settings.” “LG AI Research plans to continue identifying new areas where AI can be used to solve complex industrial challenges.”
The company is also demonstrating EXAONE Business Intelligence, a financial AI agent that analyzes approximately 8,000 publicly traded companies in South Korea and the United States daily and generates predictive scores and analyst-style comments to support investment decisions.
Following its partnership with the London Stock Exchange Group earlier this year, the company recently signed an agreement with the Korea Securities and Accounting Corporation to expand into the domestic market.
We also introduce EXAONE Data Foundry, a platform that automatically generates high-quality datasets and builds domain-specific AI models.
The company says the system increases data productivity by more than 1,000 times and improves data quality by more than 20% on average. In a pilot with the National Pension Service, the platform generated more than 10,000 specialized data entries per day.
Separately, LG AI Research will present 14 research papers at the conference. The company’s AI model for materials generation ranked second in the world in the LeMat-GenBench benchmark, which assesses the ability of AI systems to generate stable novel crystalline materials.
