Summary
AI policy is no longer just a matter of R&D subsidies. President Lee Jae-myung's pledge to turn AI's gains into opportunities for everyone signals that the government intends to bring AI down to the level of everyday life infrastructure for ordinary citizens.
For investors, there is one question that matters most. It is less about who builds the best model and more about who controls the distribution channels and data needed to convert a nationwide AI service into real traffic.
What Happened
The remarks came at an event unveiling the AI Basic Society initiative. The language sounds like social policy, but the sentence the market should be reading is industrial policy. The plan to extend AI's benefits beyond productivity gains at a handful of companies into a usage opportunity for every citizen touches public-sector demand, platform adoption, and cloud infrastructure all at once.
The government is already advancing a domestic AI service that citizens can use free of charge and without usage caps — the so-called "AI for Everyone" project. Kakao (035720) has formally confirmed it will participate, while Naver (035420) says it will decide after reviewing the request for proposals. SK Telecom (017670), LG Uplus (032640), and KT (030200) have also been named as either participants or candidates under review. The timeline matters just as much: operator selection is expected next month, a beta service in September, and a full launch targeted by year-end.
That pace could either shorten the shelf life of the policy theme or, alternatively, bring forward the point at which actual results get tested. With government projects, share prices typically move first on the announcement, then face a second test at the selection and budget-execution stages. Platform stocks tend to price in expectations first; cloud and telecom stocks react once actual traffic and infrastructure usage are confirmed.
Structural Background
AI Basic Society is the domestic-market version of sovereign AI. Reducing dependence on U.S. and Chinese models requires Korean-language data, everyday service touchpoints, and public-data governance. In this context, KakaoTalk, Naver's search and cloud businesses, and telecom carriers' subscriber bases are not simply promotional channels — they are the distribution networks that turn AI services into something people use every day.
Breaking down the supply chain in Yoon Jae-ho's framework, the order is clear: the model is the engine, the cloud is the fuel tank, and the platform is the driver's seat. Once the government rolls out a free, nationwide service, someone has to absorb the inference cost. That means the beneficiaries extend well beyond the app screen, reaching GPU servers, data centers, networks, security, and outsourced operations. That said, a free service generates costs before it generates revenue. As the user base grows, inference costs are incurred immediately, while the monetization model comes later.
Impact on Stocks and Sectors
- Kakao (035720): Its formal confirmation of participation is the most direct catalyst. Pairing KakaoTalk's reach with its in-house AI model, Kanana, could lower the cost of acquiring early users for a nationwide AI service. That said, if the free service only drives traffic without converting it into advertising or commerce revenue, the cost burden will show up first.
- NAVER (035420): Whether it will participate is still under review. Given its combined strengths in search, cloud, and generative AI, the quality of the policy benefit would be high if it is selected. Conversely, if it opts out or ends up with a limited role, some of the sovereign AI premium the market has priced in could be given back.
- SK Telecom (017670), KT (030200), LG Uplus (032640): For the telecom carriers, the significance lies more in distribution and infrastructure than in model competition. Their subscriber bases, authentication systems, customer service operations, and network data all dovetail with everyday AI services. Revenue is more likely to show up first in B2G and B2B operating contracts than in B2C subscriptions.
- Samsung Electronics (005930), SK Hynix (000660): The policy itself does not translate directly into memory orders. But if public AI services drive up inference traffic, it adds supporting logic to the demand case for server memory and HBM. What matters here is the actual scale of data center investment, not just expectations.
Bull vs. Bear Scenarios
The bull scenario is one in which the selected companies capture early traffic from the national service and expand it into advertising, commerce, and enterprise AI applications. If usability holds up in the September beta and the year-end launch is not delayed, platform companies' AI investment spending could be re-rated from a cost item into a strategic asset.
The bear scenario is one in which the free AI service remains a cost center. Unlimited usage is good for users but means an inference-cost burden for operators. Unless government budgets, cloud pricing, and model lightweighting all fall into place, rising traffic could squeeze operating profit margins. Investors should also be wary of the classic pattern in which a policy theme fizzles out once the selection news is behind it.
Investor Action Points
- Watch next month's operator-selection results to check the actual scope of Kakao's, Naver's, and the three carriers' roles — there is a real difference between having your name attached and being the actual operator.
- In the September beta, focus on the return-visit rate and the main use cases rather than subscriber counts. For everyday AI services, the value lies in repeated use, not the first try.
- Around the year-end launch, check whether the government budget, the private-sector cost-sharing structure, and the cloud procurement approach are disclosed. Who ends up bearing the cost is the second key question for share prices.
- In the next earnings season, compare the growth rates of server costs, R&D spending, and marketing expenses before looking at AI-related revenue. That is where the narrative and the income statement tend to diverge.
Kakao: A Real-Time Data Snapshot
Kakao's most recent closing price was KRW 36,200 (-2.56% versus the prior day), and the composite signal combining foreign and institutional order flow (supply-demand) with news and momentum reads 🔴 Caution. With foreign investors, institutional investors, and momentum all turning negative, caution is warranted right now.
- ▼ Dual-side selling — foreign investors and institutional investors sold a net KRW 3.0 billion and KRW 200 million, respectively, in tandem
- ▼ 52-week range position — near the 52-week low, at the 11% mark
Recent related news skews favorable, with 2 positive catalysts and 0 negative catalysts.
※ Price and foreign/institutional order-flow (supply-demand) data are provided by Korea Investment & Securities (KIS) and reflect figures as of publication time.
This article is automatically summarized and analyzed based on the original news report. View original article (Yonhap News, Industry)
