Scott Latty, senior vice president at global credit rating and research firm Morningstar DBRS, said in an interview with the SK Telecom newsroom on the 23rd that the SK Telecom consortium participating in Korea's "independent AI foundation model project" has the core capabilities to carry out the project.
In a report at the end of Nov. titled "Telecom companies are well positioned to benefit from sovereign AI infrastructure plans," Morningstar DBRS defined sovereign AI as an area that requires execution at the level of national infrastructure and identified telecom companies as the realistic driving force.
Regarding the government's selection of the SK Telecom consortium as one of the project's elite teams, Latty said it was "a strategically very meaningful choice." The intent is that the combination of SK Telecom's AI data center operating experience and AI execution capabilities, along with its know-how in reliably operating large-scale complex networks, raises the likelihood of success. He also said that being assigned the role of supporting GPUs within the elite team is a signal that SK Telecom will play a major role in providing GPU infrastructure and related services.
He assessed that telecom companies can encompass the network, data center, and power collaboration structure, making them "entities that can actually integrate and build sovereign AI infrastructure." He also saw that their experience operating infrastructure in line with each country's laws, regulations, and cultural norms, and their track record of cooperation with governments, underpins trust in the highly sensitive area of sovereign AI.
As a differentiating factor, he cited the capability to develop AI models optimized for the Korean language and culture. Given that Korea's 5G and optical communication infrastructure is conducive to the spread of data-intensive AI services, the outlook is that a telecom-led sovereign AI model can be competitive in non-English-speaking markets.
Latty emphasized that "sovereign AI will broadly change business operations, the structure of economic growth, and people's daily lives," and said it will become a core technology that will determine national competitiveness over the next 10 years. Considering its capital-intensive nature, he advised that long-term investment, along with execution capabilities and setting priorities, will be key.