SKT is expanding the generative AI market for enterprises.

SKT announced on the 23rd a 'Multi LLM (Large Language Model)' strategy to target the enterprise and public sector 'Generative AI market.' SKT stated that it is currently making technological and business preparations to provide customized solutions to customers by combining A.dot LLM, Entropic's LLM, and Conan LLM to meet the needs of enterprises and public institutions.

SKT stated that it is significant in that it “revealed plans to pursue a ‘multi-LM’ business with a lineup that includes its own LLM as well as other companies’ LLMs.”

SKT A. LLM, based on the Korean language, has been launching a basic model for enterprise since August and is training and optimizing models for client companies.

SKT plans to utilize Conan's LLM. In addition, regarding Anthropic, for which it announced an equity investment on the 13th, it is currently discussing a sales method that combines LLM functions with SKT's major B2B services. Actual service provision is expected to be possible starting from the fourth quarter when the functional integration work is completed.

SKT stated, “A ‘Full Stack Service Approach’ will be possible in the enterprise AI field, ranging from hardware infrastructure to LLM software.” In particular, it utilizes new technologies such as the X330, a new product from Sapion, an AI semiconductor specialist established by SKT, and is based on the Titan supercomputer.

SKT aims to provide 'industry-specific LLM services' for the enterprise and public sectors. SKT targets the Korean-language on-premise LLM market with A.NET LLM at the center, while leveraging tools like Entropic LLM for the cloud-based LLM market, where competition is expected to be dominated by massive AI.

Meanwhile, SKT has announced that it will build an ecosystem for promoting LLM businesses in various ways, including pursuing cooperation with Persona AI on LLM implementation projects in the AICC field to collaborate with operators who have extensive experience in actual implementation and operation.

SKT stated, “The sectors where demand for LLM services is emerging first include government ministries, public enterprises, and financial institutions,” adding, “Although it is still in the early stages, demand for AI in the manufacturing sector is also emerging, centered around major domestic manufacturers.”

SKT is preparing to participate in the LLM-based pilot and main projects currently being promoted by the government, including the Ministry of the Interior and Safety and the Ministry of Science and ICT, and is also pursuing LLM-based service construction projects centered on the financial sector together with individual clients. Discussions are also underway with group affiliates, such as SK Hynix, regarding the application of LLM AI technology to the manufacturing sector.

Kim Kyung-duk, Head of the CIC (Vice President) leading SKT’s enterprise business, emphasized to employees attending the multi-LM training course, “What matters is customer needs, and it is important to quickly provide the services that customers actually require based on various multi-LM combinations.”