What happened
Chosunbiz reports that SK Telecom, Upstage and LG AI Research Institute are competing in the final stage of South Korea’s independent AI foundation-model project. Securities analysts have highlighted the potential effects on related companies, but concerns about evaluation transparency and changing government initiatives remain unresolved.
Chosunbiz reports that three consortia—SK Telecom, Upstage and LG AI Research Institute—remain in contention for the government’s independent AI foundation-model project. The outlet says Hana Securities identified SK Telecom as a preferred telecom stock because its selection prospects were rising, while Kiwoom Securities raised its target price for LG Corp. after LG AI Research Institute’s EXAONE ranked first in the first-round evaluation. Samsung Securities and Hana Securities also cited the potential value of SK Networks’ 11.7% stake in Upstage. These are brokerage assessments, not independently verified results or guarantees of future corporate value.
The report says Motif Technologies’ consortium ranked first on the intelligence index used by Artificial Analysis in the second-round evaluation but finished last overall and was eliminated. After Motif objected, questions arose about the transparency of the evaluation categories and scoring. Chosunbiz also reports that participants are concerned about newer government initiatives, including Everyone’s AI and a frontier AI model, potentially changing project priorities and the roles originally expected of participating companies.
Source details: biz.chosun.com ↗
Why it matters
The report shows how a government-backed AI program can affect public-company valuations before its commercial results are established. That matters because investors, participating companies and policymakers may be assigning significant value to selection prospects without clear evidence that the resulting models will generate durable revenue, competitive advantage or practical adoption. Chosunbiz’s account is based on analyst and academic comments; the claims and market implications were not independently confirmed from primary government, company or evaluation documents in the supplied source.
The project has consequences beyond research funding because securities analysts are using possible selection and equity stakes to reassess listed companies. That creates a gap between policy value, technical performance and commercial value: a model may support national AI goals without producing near-term revenue for the companies involved.
The supplied report does not establish that any consortium has achieved superior real-world performance, secured broad deployment, or obtained confirmed commercial advantages. A Kyunghee University professor cited by Chosunbiz said selection alone would not automatically increase corporate value, while a Sejong University professor called for clearer evidence of tangible benefits.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
Which component of an AI application is the machine-learning model itself?
What to watch next
The key unknowns are the final selection, the government’s concrete support and access arrangements, the evaluation methodology, and whether any selected model produces measurable commercial or public-sector benefits. It is also unclear how the newer Everyone’s AI and frontier-model initiatives will affect the current project.
The final consortium decision and the government’s published evaluation criteria should clarify how technical scores, ecosystem goals and policy priorities were weighted. The source does not provide the final decision, a complete scoring framework, or a primary government record.
Investors will need evidence of actual use, customer demand, revenue, deployment conditions and support commitments before treating selection as durable corporate value. The source does not document access terms, pricing, model availability, results beyond the reported evaluation references, or confirmed financial outcomes.
The relationship between the existing foundation-model project and newer initiatives such as Everyone’s AI remains uncertain. Chosunbiz reports participant concerns that responsibilities may shift, but the source does not establish whether the programs will be merged, funded separately or given distinct roles.