The not-usually-so-contentious trade and tech relationship between the U.S. and South Korea is top of mind this week, with the convergence of several developments. First one: OpenAI opening an office in Seoul, seeking to support sovereign AI and deepen its local relationships, including with the likes of Samsung and SK. However, those same two companies are facing the likelihood of the U.S. government placing additional export regulations on their practice of putting U.S. chip-making equipment into Chinese plants (note, this is for general-purpose chips, not for AI chips). And although it’s not telecom or AI news per se, it’s hard to imagine there won’t be some kind of fallout in the U.S./Korea relationship from the recent federal immigration raid on an EV plant in Georgia being built by Korean company Hyundai, which resulted in more than 300 South Korean nationals employed at the site being detained; they are being sent home via charter flight.
South Korea had announced $350 billion in investments in the U.S. earlier this year as part of a trade deal that is still being finalized, which is supposed to result in the longtime U.S. ally receiving a tariff reduction from 25% to 15% on products from South Korea imported to the U.S. That deal is currently in tense negotiations. However that specific deal turns out, though, the global competitive landscape is shifting fast these days, and it’s unlikely to be the last word in where the lines are drawn on tech, trade and competing national interests.
Kelly Hill
Executive Editor
RCRTech
AI Infrastructure Top 3
OpenAI eyes Korean DC: OpenAI opens a Korean office and hints at building a data center, deepening ties with Samsung and SK, as Seoul advances its sovereign AI ambitions.
Chip waiver for Korea?: The U.S. continues to tighten exports of tech — in this case, potentially requiring limited yearly approvals for Samsung and SK Hynix to ship chipmaking tools to their Chinese plants.
HPE’s run at AI networking: HPE’s Juniper-fueled push puts it squarely in the AI networking race, promising self-healing enterprise systems with Mist AI and Aruba — but integration, trust, and talent gaps stand in the way.
AI-Powered Telecom Infrastructure
Supermicro, in collaboration with NVIDIA, delivers AI-powered infrastructure tailored for telcos, enhancing operational efficiency, network management, and customer experiences. Explore now
AI Today: What You Need to Know
Record-high AIDC spend: U.S. data center construction hit a record $40B in June, up 30% year-on-year, as Microsoft, Alphabet, and Amazon ramped up their AI infrastructure and boosted demand for Nvidia’s chips.
ND for DCs: North Dakota’s cool climate, surplus energy, and economic strength are seen as advantages for siting AI infrastructure there and enabling more efficient, lower-cost operations than southern states.
Digital Realty launches lab: Digital Realty launched its Innovation Lab in Northern Virginia, offering enterprises a live data center environment to test, validate, and optimize AI, HPC, and hybrid cloud deployments.
WIA on AI: AI’s future depends not just on chips and hyperscale data centers but on wireless connectivity, edge computing, and real-time infrastructure, writes WIA President and CEO Patrick Halley.
Google AI, a polyglot: As Google seeks to boost users’ AI search options, it is expanding its AI Mode to return results in additional languages: Hindi, Indonesian, Japanese, Korean, and Brazilian Portuguese.
The concept of consciousness: Humans are debating whether advanced AI is conscious. But the reality is that humans barely understand what “consciousness” actually means for their own species.
AI upskilling: OpenAI and Cisco have pledged support for massive AI upskilling, aiming to certify and train millions of U.S. workers, and signaling a corporate-led shift toward skills-based employability in the AI era.
Age of reasoning: Reasoning models surged in 2025, with enterprises rapidly adopting OpenAI o1 and DeepSeek R1. But high training and deployment costs remain a major barrier to scaling.
AI in software: AI is reshaping software engineering through code generation tools, but rising bug rates and mounting technical debt highlight the risks of rushed adoption, and the long-term complexity.
AI law firm buys law firm: Google-backed AI legal-tech platform Lawhive, whose AI colleague ‘Lawrence’ works at junior lawyer level, has acquired U.K.-based Woodstock Legal Services. It is the first time an AI platform has acquired a traditional U.K. law firm.
