The U.S. government's ban on exporting Anthropic's advanced Mythos and Fable 5 AI models is accelerating a trend it may have intended to prevent: Asian companies are now building their own frontier-level artificial intelligence systems rather than relying on American technology. Within days of the export restriction taking effect approximately two weeks before late June 2026, a Tokyo-based startup unveiled a model it claims matches Anthropic's capabilities, while a Chinese cybersecurity firm released competing tools designed for the same purpose. The developments suggest that export controls, while intended to protect American technological advantage, may instead be fragmenting the global AI market and reducing U.S. influence in key regions.

What Triggered This Wave of Asian AI Models?

The Trump administration banned Anthropic from distributing Mythos and its more restricted version, Fable 5, to customers outside the United States approximately two weeks before late June 2026. The move was framed as a national security measure, but it created an immediate vacuum in Asian markets where these models had become essential tools for businesses and government agencies. Within days, two companies stepped in to fill that gap.

Sakana AI, a Tokyo-based startup founded in 2023 by former Google engineers David Ha and Llion Jones, along with Ren Ito, a former executive at Stability AI, launched Fugu on the same week as the Mythos ban. The company explicitly positioned Fugu as a solution for organizations seeking "frontier capability without the risk of export controls." Meanwhile, Chinese cybersecurity firm 360 unveiled Tulongfeng, an AI tool designed to automatically discover software vulnerabilities, and Yitianzhen, built to automate cyber defense and incident response.

Are These New Models Actually Competitive?

Sakana AI claims that Fugu "stands shoulder-to-shoulder with leading models like Anthropic's Fable 5 and Mythos Preview." The model is designed to orchestrate access to multiple AI systems through their application programming interfaces (APIs), meaning it can coordinate between different tools rather than relying on a single provider. This architecture reflects a broader philosophy at Sakana: building AI systems that reduce dependence on any single source of technology.

The company's timing was strategic but, according to a Sakana spokesperson, not entirely planned. "Sakana Fugu is something we have been building since last year," the spokesperson explained. "The research behind it was presented at ICLR this spring, and it reflects an approach that is central to how we deliver frontier-level value at Sakana AI. We were confident in the product on its own merits; the timing simply happened to coincide with a moment that brought it more attention than we expected".

"Access to top models can disappear overnight. Collective intelligence is the practical hedge against this concentration of power," said David Ha.

David Ha, Co-founder and CEO, Sakana AI

China's approach was more direct. Zhou Hongyi, founder of 360, described vulnerability-finding AI as a national strategic asset and flagged what he called the risk of "one-way transparency," a situation in which some actors could access advanced capabilities while others could not. This framing positions China's new AI tools not just as commercial products but as essential infrastructure for national security.

How Are These Companies Justifying Their Models?

Both Sakana and 360 are making arguments that go beyond simple market competition. They are challenging the premise of U.S. export controls themselves. Sakana co-founder Ren Ito, speaking at the G7 summit in Evian, urged the U.S. federal government to "preserve access" for America's closest allies and argued that "AI should not become a technology that is hoarded; it should be one that is developed together." This position reflects a growing tension between national security concerns and the reality that AI development is increasingly global.

Sakana's approach emphasizes that the company is not trying to replace U.S. AI models but rather to hedge against the risk of sudden access loss. "U.S. models remain important to Asia," the company stated. "We'd characterize the current moment in those terms rather than as a permanent realignment toward any one set of players." However, the company is explicitly targeting Japanese businesses and government agencies looking to reduce their exposure to export controls.

What Are the Broader Implications for the AI Race?

The emergence of competitive Asian models within weeks of the Mythos ban raises questions about whether export controls can effectively slow the global AI race. Anthropic had been on a historic growth trajectory, with its run-rate revenue crossing $47 billion in May 2026. However, the company's ability to capture that revenue from Asian markets is now uncertain. Even if the U.S. government eventually lifts the ban, local alternatives trained to understand local languages and cultural nuances are already filling the gap.

The situation illustrates a fundamental challenge for U.S. policymakers: restricting access to American AI technology may protect short-term national security interests, but it also accelerates the development of competing systems in other regions. Companies like Sakana and 360 now have both the motivation and the market opportunity to build frontier-level AI models tailored to their regions.

How Are Companies Adapting to Export Control Uncertainty?

• Building Regional Alternatives: Asian startups are developing AI models optimized for local languages, cultural contexts, and regulatory environments, reducing dependence on U.S. exports and creating sustainable competitive advantages in their home markets.
• Emphasizing Orchestration Over Scale: Companies like Sakana are designing models that coordinate between multiple AI systems rather than trying to build the largest single model, reducing the risk that any single export ban could disrupt operations.
• Framing AI as National Infrastructure: Both Japanese and Chinese companies are positioning AI development as essential to national security and economic resilience, justifying government support and market protection for homegrown alternatives.

The speed at which Sakana and 360 launched their models suggests that the technical barriers to building frontier-level AI have lowered significantly. Both companies were able to develop competitive systems while the Mythos ban was still fresh, indicating that the expertise and resources needed to build advanced AI are no longer concentrated in the United States. This shift could reshape the global AI landscape for years to come.