Anthropic’s Seoul Office Opens as U.S. Orders a Halt to Its Best Models in Korea
Anthropic opened a new office in Seoul on the same day the U.S. government ordered it to stop offering its most powerful models in South Korea, a coincidence of timing that captures the company’s predicament in Asia. The company’s Claude Fable 5 and Mythos 5 models were pulled from the Korean market at Washington’s direction, according to people familiar with the matter, even as Anthropic was expanding its physical presence there. South Korea has been one of Anthropic’s most important Asian markets, and the episode shows how export controls are shrinking the territory the company can serve.
The order is part of a broader pattern of U.S. restrictions on advanced AI models, which Washington treats as sensitive technology that should not be made available everywhere. Anthropic has generally complied, positioning itself as a company that welcomes oversight and works with regulators. But compliance has a cost: every market that Washington closes is a market where competitors, including Chinese developers and open-source alternatives, can gain ground. The Seoul episode is the most visible example yet of the tension between the company’s expansion ambitions and the constraints placed on it from Washington.
The timing of the office opening made the tension literal. Anthropic executives attended a ribbon-cutting in Seoul, greeted local partners and employees, and discussed plans to grow the company’s business across the region, all while the models that Korean customers had come to rely on were being switched off. A person close to the company said the two events were handled separately, with the office opening proceeding on its own schedule, but the message to the Korean market was unavoidable: Anthropic wants to be in South Korea, and Washington is deciding what it can sell there.
The company’s troubles in Asia extend beyond Korea. Anthropic’s service limitations in India have drawn criticism from Indian media, which have questioned whether American AI companies are sacrificing emerging markets as the U.S.-China technology competition intensifies. India has been one of the largest markets for AI usage, with developers and businesses adopting models at a rapid pace, and restrictions on access to leading American models have pushed many of them toward open-source alternatives that can be hosted locally. The criticism has a political edge, framing the restrictions as the United States using its control over AI to constrain the growth of partner economies.
Anthropic’s response to the criticism has been to argue that the restrictions are government policy, not company preference, and that it is working to provide alternatives within the rules. The company has pointed to its efforts to offer models in markets where it is permitted, and to its engagement with regulators in multiple countries. But the distinction between what Anthropic wants and what it is allowed to do is lost on many users, who see only a leading AI provider withdrawing from their market while continuing to operate in the United States and Europe.
The broader strategic question is whether the compliance-first approach still works. Anthropic has built its reputation partly on its willingness to engage with regulators, requesting oversight in the United States and presenting itself as the responsible alternative in an industry full of wilder competitors. That posture has won it credibility in Washington, but it also makes the company the easiest target for export controls, since it can be counted on to comply. Rivals with less cooperative reputations face the same restrictions but generate more friction, and some market observers argue that friction is itself a form of competition.
For India and South Korea, the episodes are accelerating a shift that was already underway. Both countries are investing in domestic AI development and have seen a surge in adoption of open-source models that run without American permission. The restrictions have given that shift a rationale: if American models can disappear at Washington’s discretion, dependence on them is a strategic risk. Governments that once treated American AI as infrastructure now treat it as a resource that may not always be available.
Anthropic’s path forward, people familiar with its thinking said, is to comply with the current rules while working to expand the set of markets where it can operate, using its regulatory engagement as a credential for trust. It is a patient strategy, and it leaves the company’s near-term growth dependent on decisions made in Washington. The Seoul office will serve the markets it can, but its first week has demonstrated how far the company’s reach extends beyond its control. South Korea’s reaction will matter beyond the company’s own fortunes. The country has built one of the world’s most sophisticated technology sectors, with domestic champions in semiconductors, electronics and internet services, and its developers are heavy users of advanced AI models for coding, research and product development. Korean companies have been investing in their own large language models, and the gap left by Anthropic’s withdrawal could accelerate those efforts. Local analysts said the episode has already influenced procurement decisions at Korean firms, which now weigh the risk of service interruptions alongside the quality of the models when choosing an AI provider. The same calculation is being made in Japan and across Southeast Asia, where the reliability of access is becoming a factor as important as capability.
This article was prepared by Rhino Finance’s editorial team based on public reporting.
