The plan presented at the seminar
SK Group intends to adopt a strategy it calls the “AI flywheel,” which links the group’s activities in semiconductors, infrastructure and energy to new artificial intelligence services. The company reported this on Sunday, according to The Korea Herald. Chairman Chey Tae-won presented the plan at the 2026 CEO Seminar, held from Thursday to Saturday at the SKMS Research Institute in Icheon, Gyeonggi Province. About 50 executives attended the meeting, including Executive Vice President Chey Jae-won and SK Supex Council Chairman Chey Chang-won.
According to the source’s account, Chey Tae-won asked the affiliates to break out of existing business models and redefine themselves as AI companies, identifying opportunities beyond traditional activities. He also stressed that in the AI era no one can do everything alone and that open collaboration with global partners is essential to strengthen capabilities and accelerate progress.
How the “flywheel” would work
The mechanism described calls for SK’s AI chips to support the expansion of so-called AI factories, driving up demand for electricity and for energy infrastructure and services. Growth in these sectors should in turn fuel further investment and innovation in the semiconductor sector, in a cycle that justifies the term “flywheel.”
The group also aims to expand its presence in artificial intelligence models and applications and to take part in what it calls the “token economy,” namely the exchange of AI computing resources and services. This is a strategic indication: the material does not specify which products, timelines or investments are envisaged for this expansion.
During the seminar, executives examined emerging opportunities and potential risks, agreeing to continue the portfolio rebalancing initiative called Portfolio Rebalancing 2.0, aimed at streamlining inefficient activities and strengthening core ones. The group also plans to extend AI-driven transformation to its own operations to improve productivity and profitability.
The applications already cited and the open points
The material cites some ongoing or announced applications. SK hynix is upgrading its smart factories with AI and digital twin technology to improve production efficiency. SK Innovation has introduced an ontology-based process map that links plant and operational data to improve production yields at the Ulsan CLX complex. SK Telecom is automating network and service operations to increase productivity and customer experience, while SK Gas uses AI to analyze market conditions and organize operational knowledge to support faster decisions.
The significance of the plan lies in the attempt to link traditionally separate businesses — chips, energy and infrastructure — into a single AI-related demand logic, with an explicit positioning toward models, applications and computing services. Several elements, however, remain to be verified: the material does not indicate amounts, deadlines or measurable targets for the “AI flywheel” or for entry into the “token economy,” and it does not clarify which global partners are involved. Nor are the criteria specified for how the portfolio rebalancing will affect individual subsidiaries. The available information comes from the company statement reported by the source.
