SK Telecom Co Ltd

Three consortiums led by SK Telecom, KT and Kakao plan to launch artificial intelligence services developed under the South Korean government's project in closed beta within October. Under the AI for All project, launched last July, the initiative aims to reduce reliance on foreign AI models and requires developers to use domestically developed AI models for at least 50% of the services they offer, while receiving government funding and GPU support. Initial developers are targeting 5 million monthly users, while Kakao aims to grow that to 10 million by 2027 and 30 million by 2030. The government also plans to inject an additional 250 billion won in budget starting next year, part of which will be used to procure around 2,000 GPUs. Meanwhile, the Ministry of Science and ICT said that about 23 million South Koreans have used generative AI models, with roughly 21 million using free basic-tier models.

Samsung Electronics signed AI radio-network deals with KT and SK Telecom, even as its Seoul shares closed Monday down 5.43% at 270,000 won. The trials begin in October, with KT planning to test welding robots at a shipyard and SK Telecom testing patrol robots and video monitoring at a petrochemical site. Samsung is KT's sole global vendor and SK Telecom's main vendor for the projects, but the company disclosed no contract value, making the trials a proof point rather than a revenue forecast. Samsung's GF Score stands at 84 out of 100, showing strong growth and financial strength alongside a much weaker GF Value reading. Investors now need to see whether successful trials lead to repeat deployments and contracts large enough to matter.

South Korea's Ministry of Science and ICT (MSIT) announced today that it has selected three private consortiums, led by SK Telecom, KT Corp, and Kakao Corp, to jointly develop artificial intelligence (AI) services accessible to all citizens under the "AI for All" project. Each consortium will receive a total of 512 Nvidia B200 graphics processing units (GPUs), and the government will fully fund the launch and operation of these services next year. The ministry plans to sign agreements with each consortium in September and aims to officially launch the services by the end of this year. The project, initiated in July, aims to drive the country toward an AI-based society, requiring developers to use domestic AI models for at least 50% of the system. A total of six consortiums applied for the competition.

SK Telecom announced it will spin off its subsidiary SK Broadband into a surviving company and a new AI data center company, SK Horizon, securing a combined KRW 3.08 trillion equity investment from KKR and the IMM Investment-Stonebridge consortium. The spin-off ratio is set at approximately 0.84 for the surviving company and 0.16 for the new company. SK Horizon will manage eight existing data centers and new ones under construction, targeting a total capacity of 318 MW, while SK Hyper will develop GW-scale projects with a goal of 5 GW by 2029 and 15 GW by 2035. KKR and IMM will hold 29% and 20% stakes respectively, with SKT retaining 51% and management control. The spin-off is expected to complete in the first quarter of next year, subject to shareholder and regulatory approvals.

Anthropic, the AI lab founded in 2021 by former OpenAI employees, is set to become the most anticipated IPO of 2026 after filing its Form S-1 with the SEC on June 1st, with Polymarket betting showing an 83% chance of a public debut by the end of October and 93% by year-end. With a private valuation of approximately $965 billion and an annualized revenue run rate of about $65 billion, Anthropic is the fastest-growing technology company in history, specializing in enterprise AI and software coding through its Claude models. Several public companies hold significant stakes, including Amazon, whose $43 billion stake is only about 2% of its market cap, while SK Telecom's $2.7 billion stake represents roughly 20% of its market cap, making it the biggest relative beneficiary. The IPO is seen as a crucial barometer for the AI industry, given that AI spending accounts for the lion's share of US GDP growth.

BitGo Korea, the South Korean subsidiary of US digital asset infrastructure company BitGo, announced on August 20 that its registration as a crypto asset business was accepted by the Korea Financial Intelligence Unit under the Financial Services Commission, effective August 18. According to South Korean English-language daily The Korea Herald, this is the first time an overseas crypto asset company has established a new entity in South Korea and received registration acceptance as a VASP. BitGo Korea is approximately 25% owned by South Korean financial group Hana Financial Group and about 10% by telecom giant SK Telecom, and plans to fully launch custody, management, and transfer services for crypto assets targeting South Korean financial institutions and corporations.

SK Telecom reported a sharp rise in second-quarter net income, with profit attributable to shareholders of the parent company climbing to KRW 470.6 billion from KRW 89.6 billion a year earlier. Operating income increased to KRW 566 billion from KRW 338.3 billion, while sales edged up to KRW 4.36 trillion from KRW 4.34 trillion. Shares of the South Korean wireless operator were trading 2.30% higher at KRW 93,500 on the Korean Stock Exchange.

Nvidia and South Korea's SK Group have expanded their collaboration with an initiative worth over $500 billion spanning AI data centers and next-generation memory. The partnership includes plans for SK Telecom to build a 2-gigawatt-scale AI Cloud in South Korea using the Nvidia DSX platform and Vera Rubin accelerated computing powered by SK hynix HBM4 chips, with the first AI factory expected online in 2027. SK hynix is also entering a long-term AI memory partnership with Nvidia to secure stable supply and co-develop next-generation solutions including High Bandwidth Memory. The deals are part of $750 billion in agreements signed by SK Group and contribute to South Korea's broader $950 billion in new AI initiatives involving Samsung, SK Group, and U.S. tech companies. Nvidia separately plans to invest $1 billion for a 4.5% stake in Naver, while Samsung has inked a memorandum of understanding worth more than $200 billion to supply Broadcom with memory and foundry chips through 2030.

SK Group and NVIDIA have announced a $500-billion-plus comprehensive partnership to establish AI infrastructure serving surging global compute demand. The collaboration includes SK Telecom building a 2-gigawatt NVIDIA Vera Rubin DSX AI factory in Korea, with the first facility expected online in 2027, and a long-term AI memory partnership between SK hynix and NVIDIA to codevelop and secure next-generation AI memory including HBM. The AI factory will use the NVIDIA DSX full-stack architecture and SK hynix HBM4 to deliver low token cost at maximum energy efficiency, targeting sovereign, physical, agentic, and enterprise AI services across the Asia-Pacific region. SK Group Chairman Chey Tae-won said the partnership will help Korea become a global hub driving AI innovation, while NVIDIA founder and CEO Jensen Huang noted that South Korea has all the ingredients to become a global AI powerhouse.

SK Telecom announced plans to build an AI data center with a capacity of up to 15GW, aiming to make Korea a leading AI infrastructure hub in Asia. The project aligns with the government's 'AI G3' strategy to position Korea among the top three AI powers alongside the U.S. and China. The Ulsan AI Data Center will expand to GW scale, with 5GW to be activated in stages starting in 2029 and ultimately reaching 15GW. A 1GW-class AI data center could cost around KRW 70 trillion, with financing from SKT's own investment, strategic partners, long-term customer contracts, and project financing. SKT also plans to build a 2GW cluster in the southeastern Gyeongsang region and an additional 1GW in the southwestern Jeolla region, bringing domestic capacity to 5GW by 2029.

A bullish thesis on Penguin Solutions, Inc. highlights its transformation into an AI infrastructure platform centered on high-performance computing, enterprise AI deployments, and next-generation memory architecture. The company operates through Advanced Computing, Integrated Memory, and Optimized LED segments, with Advanced Computing shifting from volatile hyperscaler contracts toward enterprise, sovereign AI, and government customers. Integrated Memory benefits from AI-driven DRAM demand and DDR5 upgrades, while CXL-based MemoryAI products address AI inference bottlenecks. A strategic partnership with SK Telecom and SK Hynix provides capital and potential access to advanced memory technologies. The thesis also notes potential upside from a future divestiture of the LED business and growth in recurring software revenue through the ICE ClusterWare platform.

Nvidia CEO Jensen Huang announced strategic artificial intelligence partnerships with six major South Korean companies following a four-day visit. The collaborations span advanced memory chips, humanoid robots, self-driving cars, and massive AI data centers. SK Hynix remains Nvidia's largest memory partner, with procurement expected to rise well above current multi-billion-dollar levels. SK Telecom plans to build a gigawatt-scale AI cloud using Nvidia technology by 2027, while a partnership with LG Group focuses on humanoid robotics and data center infrastructure. Nvidia also intends to build an AI Valley data center in Saemangeum and expand autonomous mobility and AI manufacturing operations after discussions with Hyundai Motor Group Executive Chairman Euisun Chung.