Days before Anthropic took its most advanced AI models offline, the White House directed the company to revoke SK Telecom’s access to the Claude Mythos system. The order stemmed from allegations concerning potential ties between the South Korean telecom giant and China.

Background of the Controversy

SK Telecom is a leading telecommunications provider in South Korea and a significant investor in Anthropic. The company had been granted access to Claude Mythos, a highly advanced AI model developed by Anthropic.

The White House intervention followed intelligence assessments that raised questions about SK Telecom’s business relationships and data-sharing practices with Chinese entities. No specific evidence of wrongdoing has been publicly disclosed.

Implications for AI Governance

This incident highlights the growing tension between national security concerns and international collaboration in AI development. The U.S. government has increasingly scrutinized foreign access to frontier AI systems, particularly those with potential dual-use applications.

Anthropic complied with the directive and subsequently took Claude Mythos offline for unspecified reasons. The company has not commented on whether these two events are directly linked.

Reactions from Stakeholders

SK Telecom issued a statement denying any improper ties to China and expressed disappointment with the decision. The company emphasized its compliance with all applicable laws and regulations in South Korea and abroad.

Industry observers note that this case could set a precedent for how AI companies manage access permissions for international partners and investors. The lack of transparent criteria for such security assessments has drawn criticism from privacy advocates.

Technical and Legal Dimensions

Claude Mythos represents a class of large language models trained on extensive datasets. Its removal has affected research collaborations that depended on its capabilities.

Legal experts point out that the White House order relied on executive authority rather than formal legislation, raising questions about the legal framework for regulating AI access in cross-border contexts.

Next Steps

Anthropic is expected to release further details about the timeline for reinstating Claude Mythos or its replacement models. SK Telecom may seek clarification through diplomatic channels or legal review. Broader policy discussions on AI export controls are likely to intensify in the coming months.