Anthropic opens Seoul office amid US export controls on its top AI models

Anthropic opened its Seoul office Tuesday with enterprise deals including Naver, Samsung SDS, LG CNS, and Nexon, and signed an AI safety MOU with South Korea’s Ministry of Science and ICT.anthropic
The launch comes days after the U.S. Commerce Department ordered Anthropic to disable its Fable 5 and Mythos 5 models globally, citing national security concerns in the first use of 2018 export control powers on a frontier AI system.reuters
Available models like Claude Opus 4.8 remain accessible, and Korean partners said they continue building on those tools as Anthropic expands its third Asia-Pacific office.anthropic

Anthropic Opens Seoul Office and Signs AI Safety Agreement With South Korea Amid U.S. Export Turmoil

Anthropic opened its Seoul office on Tuesday and announced a memorandum of understanding with South Korea’s Ministry of Science and ICT to cooperate on AI safety, even as the company’s most advanced models remain disabled worldwide under a U.S. government export control order issued just days earlier.

Korean Partnerships and Safety Accord

The office launch, announced June 17, came with a suite of enterprise partnerships spanning some of South Korea’s largest technology companies. Naver has deployed Claude Code across its entire engineering organization, with thousands of engineers using the tool for coding productivity. Samsung SDS is rolling out Claude to employees across Samsung Electronics for knowledge work and software development. LG CNS is deploying the AI assistant to thousands of employees, while Nexon engineering teams use it to write and review code for live-service games.aiweekly

Under the MOU with the Ministry of Science and ICT, Anthropic and South Korea will collaborate on evaluating model safety in Korean with the Korea AI Safety Institute and exchange information on AI-enabled cyber threats. The company also announced academic partnerships with the National AI Research Lab consortium — spanning KAIST, Korea University, Yonsei University, and POSTECH — providing Claude access to up to 60 researchers working on AI safety and alignment.anthropic

KiYoung Choi, a veteran technology executive appointed as Anthropic’s Representative Director for Korea in May, said Korean organizations “understand that innovation and safety are two sides of the same coin”.anthropic

Export Controls Cast a Shadow

The Seoul expansion arrives under unusual circumstances. On June 12, the U.S. Commerce Department issued an export control directive ordering Anthropic to suspend all access to its most capable models — Fable 5 and Mythos 5, launched just three days prior — for any foreign national, regardless of location. Unable to screen its user base in real time, Anthropic disabled the models globally within 90 minutes.fortune

The directive, sent by Commerce Secretary Howard Lutnick to CEO Dario Amodei, cited national security concerns related to cybersecurity capabilities, according to Reuters. It marked the first time the Commerce Department invoked the 2018 Export Control Reform Act to restrict access to a commercially deployed frontier AI system. Other Anthropic models, including Claude Opus 4.8, remain accessible.reuters

A Dual Signal

The juxtaposition is stark: Anthropic is deepening its presence in a key Asian market while its most powerful technology is under lock by Washington. The Seoul office is Anthropic’s third in the Asia-Pacific region, following Tokyo and Bengaluru. Korea ranks among the top dozen countries globally for Claude usage, according to Anthropic, and the company’s Korean partners said they continue to build on the models that remain available.koreaherald

The situation underscores the tension facing U.S. AI companies navigating both aggressive international expansion and an increasingly interventionist federal posture on frontier model distribution.