South Korea is moving to position itself as an exporter of "intelligence" rather than just the chips and equipment that go into building it, with SK Group chairman Chey Tae-won outlining an AI data center plan expected to involve more than KRW1,000 trillion (approx. US$652.7 billion) in investment, according to Hankyung.
AI demand and capacity crowd-out effects are driving higher prices and volumes for high-voltage products, according to IC distributors, who say early pull-ins and price negotiations have become the market norm. But as concerns over an AI bubble resurface, some industry players warn that if a profitable AI business model does not emerge soon, the sector may not even make it to the ninth inning.
AI and robotics are moving from pilot projects to factory floors worldwide, but adoption remains uneven. Humanoid robots draw the headlines, yet most manufacturers still favor task-specific tools, digital twins, and collaborative machines that promise steadier gains in efficiency, safety, and precision across global supply chains.
South Korean conglomerates will invest a combined KRW312 trillion (approx. US$203.6 billion) in the Yeongnam region, as the government moves to turn the country's southeast into a hub for advanced manufacturing, AI infrastructure, semiconductors, aerospace, defense, and energy.
SEMI has warned the Trump administration that intervening in memory-chip pricing or production capacity could worsen a historic supply shortage driven by the artificial intelligence boom.
Amazon Web Services has told its server supply chain partners to raise shipment volumes for the third quarter of 2026, according to sources in the AI server supply chain.
The rapid expansion of AI applications is redefining what device makers need from display technology.
Test interface supplier Chunghwa Precision Test Tech. Co., Ltd. (CHPT) reported its June 2026 revenue, marking its sixth consecutive monthly revenue record as demand from the market remained strong. The company also posted record quarterly revenue in both the first and second quarters of 2026, underscoring its sustained growth momentum.
At a late-June industry forum, experts from Taiwan's TPIsoftware, the Institute for Information Industry (III), and Phison Electronics agreed that although AI is now indispensable for manufacturing, scaling it up depends less on raw model capability than on whether companies can actually trust it in operation. Fragmented data, weak governance, and cybersecurity concerns remain the primary hurdles keeping companies from moving past pilot projects into full adoption.
SoftBank Group (SBG) and telecom subsidiary SoftBank said they will set up a new company in the US in July 2026 to rent out AI computing resources, aiming to challenge CoreWeave and Nebius in the fast-growing AI cloud market. The new unit, SB Neo, is scheduled to begin operations in fiscal 2027 (April 2027 to March 2028).
