SK Telecom announced that the SKT Consortium has been selected as a core team for the Ministry of Science and ICT´s Proprietary AI Foundation Model project. The consortium brings together industry players, chip designers and leading research institutions to deliver a full-stack artificial intelligence platform that spans semiconductors, models, data and services. Members named in the announcement include Krafton, 42dot, Rebellions, Liner, Selectstar, professors from Seoul national university and KAIST, plus external researchers from the University of Wisconsin-Madison.
The consortium intends to build next-generation large-scale models with omni-modal capabilities that process text, images, speech and video. SK Telecom will contribute its in-house large language model work, including the A.X series developed since 2018 and its TITAN supercomputer training environment. The company has already open-sourced multiple models, including A.X 3.1 and A.X 4.0 variants, and documented outputs such as research papers, patents and open projects. Rebellions will supply domestically developed neural processing units to optimize inference performance and efficiency, while other partners will supply data reliability, mobility and application expertise.
Technical ambitions are paired with an explicit open-source aim: the consortium plans to release components to the domestic ecosystem to accelerate adoption by Korean companies and researchers. SK Telecom highlighted past model performance, citing parity-class results against global benchmarks in Korean language tasks, and expects to leverage its GPU and data center investments to provide large-scale computing independently for the program. The project targets practical sector use cases in office productivity, manufacturing, automotive, gaming and robotics, with an emphasis on making agent-like services accessible across Korea.
Leaders framed the selection as a step toward sovereign capability and broader industry impact. SK Telecom quantified its AI research footprint and stressed operational experience as advantage in building national infrastructure. The company said it will pursue model development, infrastructure scaling and open collaboration while aligning efforts with government goals for national competitiveness in artificial intelligence.
