For practitioners, the EXAONE demonstrations highlight how large proprietary LLMs are being applied beyond chat and code to accelerate materials discovery, financial analytics, and data generation workflows, raising practical questions about integration, validation, and IP around model-accelerated R&D. Reporting by multiple Korean outlets shows that at ICML 2026 in Seoul, LG AI Research showcased its proprietary large model, EXAONE, and three commercial-facing platforms: EXAONE Discovery, EXAONE BI, and EXAONE Data Foundry. Korea Herald and BusinessKorea report that EXAONE Discovery identified a hair-loss candidate described as "Rhamsydil" from more than 420,000 candidates in a single day and that an immersion cooling fluid for data centers was co-developed with GS Caltex. Reporting by Korea Herald and Asiae states that EXAONE BI analyzes roughly 8,000 Korean and US stocks daily and that LG has commercial ties with the London Stock Exchange Group and Koscom. BusinessKorea and Asiae additionally report patent filings and conference papers tied to the work.
LG AI Research showcases real-world EXAONE AI applications at ICML 2026