The White House identified SK Telecom -- South Korea's largest wireless carrier and an Anthropic investor since 2023 -- as a company suspected of having ties to China, and asked Anthropic to revoke SK Telecom's access to the restricted Claude Mythos model. Anthropic complied immediately, underscoring how quickly model access can become entangled in national-security policy.
The episode is a vivid example of AI's geopoliticization. Frontier models are increasingly treated as strategic assets subject to export-control logic, and the U.S. government is willing to reach into commercial relationships -- even ones involving an investor and partner -- to restrict who can use the most capable systems. For model providers, that means customer and investor relationships now carry geopolitical risk that can override commercial intent.
“Frontier models are increasingly treated as strategic assets subject to export-control logic, and the U.S.”
The move came as Anthropic was simultaneously expanding aggressively in South Korea, opening a Seoul office and signing enterprise deployments. The juxtaposition -- a major market push alongside a government-ordered access revocation -- captures the tightrope frontier labs now walk between global commercial ambition and great-power politics.