How a South Korean Telecom Partner Triggered Washington's Unprecedented Shutdown of Anthropic's Top AI Models

On June 12, 2026, the Trump administration issued an unprecedented export control order forcing Anthropic to shut down its newly released Claude Fable 5 and Mythos 5 AI models for all foreign nationals — a decision triggered by concerns that Anthropic had given early access to a South Korean telecom firm, identified as SK Telecom (SKT), that the White House suspected of having ties to China. The order sent shockwaves through South Korea's telecom industry, sparked emergency negotiations between Anthropic and U.S. officials, and raised fundamental questions about the weaponization of frontier AI access as a tool of geopolitical control .

What Triggered the Order

The direct trigger was a Washington Post report (later corroborated by multiple outlets) stating that the Trump administration's export controls were set in motion after Anthropic gave early, pre-release access to Fable 5 and Mythos 5 to SK Telecom — a South Korean telecommunications company that U.S. officials suspected of having ties to China . The White House believed a "China-linked entity" may have accessed the frontier models, raising fears of model distillation (copying the model's capabilities) and technology transfer to China .

A secondary trigger was a "jailbreak" vulnerability discovered in Fable 5 shortly after its release. Anthropic's CEO Dario Amodei acknowledged that the issue existed in "the vast majority of AI models released." A senior Trump administration official described Anthropic's handling of the issue as "recklessness" that contributed to the decision .

Commerce Secretary Howard Lutnick communicated the directive to CEO Dario Amodei via letter on June 12, citing national security authorities . The directive arrived at 5:21 PM ET, giving Anthropic no time to implement a narrower restriction . The company concluded the only way to comply was to abruptly disable both models for all customers globally, including U.S. citizens .

SK Telecom: The South Korean Firm at the Center

Multiple South Korean and international media outlets, citing the Washington Post and government officials, confirmed that the "South Korean telecommunications company suspected of ties to China" at the center of the dispute is SK Telecom (SKT) . According to reporting from the Washington Post, Anthropic had originally submitted a list of 111 organizations to receive early access to Mythos 5 as part of Project Glasswing, but later disclosed that roughly 50 additional entities had already received access — one of which was SK Telecom .

SK Telecom vehemently denied any connection to China. In a statement reported by the Kyunghyang Shinmun, SKT said: "We have no connection with China at all" . The company argued it was being unfairly implicated based on suspicion rather than evidence.

Fallout Across South Korea's Telecom Industry

The revelation put the entire South Korean telecom industry on alert, with industry players fearing guilt by association and a chill on AI partnerships with U.S. companies . South Korean media reported that the controversy highlighted how South Korean firms operating AI partnerships with U.S. companies now face heightened scrutiny over their China business relationships . Some analysts warned this could deter South Korean telecoms and tech firms from seeking early access to cutting-edge U.S. AI models.

The Korea Times reported that the episode underscored how "cutting-edge AI tools have effectively become strategic assets subject to government export controls," prompting calls in Seoul for greater "AI sovereignty" to reduce dependence on U.S. models .

Ongoing Negotiations Between Anthropic and the Trump Administration

Anthropic immediately dispatched senior technical staff to Washington over the weekend of June 13-14 for emergency negotiations . The Wall Street Journal reported that Anthropic was "racing to resolve" the dispute and engaging intensively with administration officials to lift the restrictions .

The talks centered on demonstrating that Anthropic can implement technical safeguards — such as real-time nationality verification and model access controls — that would satisfy the government's national security concerns without requiring a total shutdown . A senior Trump administration official told Fox Business that there were "ongoing talks about fixing the vulnerabilities with Fable 5," but the administration maintained that Anthropic's "recklessness" justified the action .

In a notable parallel move, Anthropic opened a new Seoul office on June 17 and publicly stated it expects the export curbs to be lifted soon, signaling confidence that a resolution is within reach . Anthropic's executives characterized the "jailbreak" issue as a common industry problem, downplaying its severity . However, early reports indicated that talks ended without the restrictions being lifted, with the government still seeking resolution of the jailbreak concerns first .

Broader Implications for International Trust and AI Access

AI as a strategic asset: The order marks the first time the U.S. government has used export controls to bar foreign nationals — including foreign employees of the company itself — from accessing a commercial AI model. This sets a precedent that frontier AI capabilities are treated like nuclear technology or advanced semiconductors .

Trust erosion with allies: The decision directly targeted a South Korean firm (a key U.S. ally in the Indo-Pacific), causing diplomatic unease. Seoul-based analysts argued the episode could push allied nations to accelerate domestic AI development, fearing that even trusted partners can be cut off from American AI tools at any time .

Global AI fragmentation: Experts cited in multiple reports warned that this kind of unilateral access restriction will accelerate the balkanization of AI — with the U.S., China, and allied blocs each developing separate AI ecosystems behind national firewalls . The U.S. government also indicated it would not grant exceptions even to G7 member countries, deepening concerns about a fractured global AI landscape .

Impact on Anthropic's business model: Anthropic, which had explicitly called for AI regulation in the past, now finds itself hurt by regulation that went "much further" than it anticipated . The company's ability to sell enterprise AI services to international customers — including allies — is now in doubt unless a negotiated framework is reached.

---

Bottom line: The June 12 order was directly triggered by Anthropic's pre-release sharing with SK Telecom, which the White House viewed as a China-link risk. SK Telecom denies any China ties. The fallout has roiled South Korea's telecom sector, Anthropic is in emergency talks to restore access, and the broader implications point toward a fractured global AI landscape where trust — even among allies — is no longer assumed.