Lee Unveils Korea AI Cooperation Declaration
South Korean President Lee Jae Myung unveiled the San Francisco AI Declaration at an AI summit on July 24, presenting Korea as a prospective global hub for AI production, deployment and cooperation. Reuters reported that Samsung Electronics and SK Group announced AI-related agreements with U.S. companies totaling $950 billion, including semiconductor, infrastructure and packaging arrangements.
South Korean President Lee Jae Myung unveiled the San Francisco AI Declaration at an AI summit in San Francisco on July 24, framing artificial intelligence as a new opportunity for Korea and calling for international technological cooperation. Reuters reported that the summit was accompanied by announced AI-related agreements involving Samsung Electronics, SK Group and U.S. technology companies totaling $950 billion.
Lee met separately with Nvidia CEO Jensen Huang, OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei and Broadcom CEO Hock Tan before the summit, according to Reuters. The Korea Herald reported that the declaration described Korea as a place where AI can be built, supplied, deployed and expanded into new markets, with human-centered development as a stated principle.
"I'd like to present a vision for the future of AI and propose global cooperation to make it happen," Lee said at the summit, according to Reuters.
Semiconductor agreements dominate the announcements
Reuters reported that SK Group signed agreements worth $750 billion, including an SK Hynix partnership with Nvidia valued at more than $500 billion. Samsung Electronics separately announced a memorandum of understanding with Broadcom covering up to $200 billion in memory chips, foundry services and advanced packaging, Reuters reported.
The announcements arrive as AI developers and cloud providers seek additional access to advanced chips and compute infrastructure. Reuters characterized the agreements as part of a broader rush by AI companies to secure supply while demand for faster chips and computing capacity exceeds supply.
Nvidia reported that NAVER, Brookfield and Nvidia would expand NAVER's NVIDIA DSX AI factory at the GAK Sejong data center to 200 megawatts, or roughly 100,000 GPUs, using the Nvidia Vera Rubin platform. Nvidia also reported a new joint AI research lab with the Korea Advanced Institute of Science and Technology, or KAIST, focused on agentic AI.
What the summit establishes
The event brought together leading U.S. AI developers with Korean firms central to the semiconductor supply chain. In remarks at a San Francisco meeting with Korean expatriates, Lee cited Samsung Electronics and SK Hynix's AI-chip market positions and argued that state-level preparation was needed as AI becomes more central to the physical world, according to Chosun.
For ML infrastructure teams, the announced agreements reinforce a practical industry pattern: advanced-model deployment increasingly depends on coordinated access to memory, packaging, foundry capacity, data-center power and accelerator systems, rather than GPUs alone. The stated values describe announced deal totals and memorandum coverage.
The declaration itself does not provide technical specifications, implementation milestones or procurement terms. Practitioner attention will therefore center on disclosed capacity, delivery schedules, hardware configurations and the availability of the research and AI-factory programs referenced during the summit.
Key Points
- 1Lee's declaration links Korea's AI agenda to international cooperation, while summit announcements center on chips, infrastructure and industrial supply relationships.
- 2Reuters reported $950 billion in announced Samsung and SK Group AI agreements, illustrating the scale of competition for advanced semiconductor capacity.
- 3Comparable AI infrastructure programs show that memory, packaging, power and data-center capacity increasingly constrain deployment alongside accelerator availability.
Scoring Rationale
The summit combines a national AI cooperation initiative with exceptionally large announced semiconductor and infrastructure agreements involving Samsung, SK Group, Nvidia and Broadcom. It matters to ML practitioners because compute availability depends on the memory, packaging and data-center supply chain, though the announcements provide limited detail on delivery timing and accessible capacity.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
