As the Trump Administration's export restrictions on advanced American artificial intelligence models tighten access for global users, companies across Asia are racing to develop homegrown alternatives—a shift that underscores how trade barriers can accelerate technological fragmentation and reshape global markets in ways policymakers may not have anticipated.
Two weeks ago, the administration banned global access to Anthropic's Mythos and its more restricted version, Fable 5. Within days, competing AI models designed to circumvent those restrictions began emerging from Tokyo and Beijing, signaling that the strategy of controlling access through export controls may be creating the very technological decoupling it aims to prevent.
The Alternatives Taking Shape
Sakana AI, a Tokyo-based startup co-founded in 2023 by former Google researchers Ren Ito, Llion Jones and David Ha, launched Fugu, a frontier AI model the company says "stands shoulder-to-shoulder with leading models like Anthropic's Fable 5 and Mythos Preview." The model is explicitly designed to help Japanese businesses and government agencies reduce exposure to tightening export controls.
The company's website advertises the model as "delivering frontier capability without the risk of export controls." A Sakana AI spokesperson said the release was "entirely coincidental," noting that research behind Fugu was presented at ICLR this spring and reflects work begun last year. The company is optimizing the model for the Japanese language and culture, targeting Japanese businesses and government agencies seeking alternatives to U.S. technology.
Meanwhile, Chinese cybersecurity firm 360 unveiled Tulongfeng, an AI tool designed to automatically discover software vulnerabilities, and Yitianzhen, built to automate cyber defence and incident response. According to Reuters, 360's founder Zhou Hongyi 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 vulnerability-detection capabilities while others could not.
The Broader Implications for Tech Access and Power
The timing reveals a structural problem embedded in export-control strategies: they create immediate incentives for competitors to build alternatives, potentially accelerating the technological fragmentation that export controls were designed to prevent. Anthropic, which saw its run-rate revenue cross $47 billion in May 2026, now faces rivals in markets where it once dominated.
David Ha, Sakana's co-founder and CEO, framed the issue in terms of systemic risk. "Access to top models can disappear overnight," he wrote on X. "Collective intelligence is the practical hedge against this concentration of power." He positioned Fugu as designed to coordinate agent usage among many models—a strategy that hedges against reliance on any single provider for critical national infrastructure.
Ren Ito, Sakana's co-founder, took the argument further at the G7 summit in Evian last week, where AI access and export controls were central topics. In an op-ed published in Project Syndicate, Ito urged the U.S. federal government to consider that its "first priority should be to preserve access" for America's closest allies. "AI should not become a technology that is hoarded; it should be one that is developed together," he wrote.
Sakana's positioning differs markedly from China's approach. While Sakana framed Fugu as a hedge strategy to preserve access to frontier AI rather than replace it, a Sakana spokesperson emphasized that "U.S. models remain important to Asia," characterizing the current moment as something other than "a permanent realignment toward any one set of players." This reflects Ito's remarks at the G7, where he stressed that export controls should not fracture the international AI ecosystem.
China's 360, by contrast, positioned its tools as national strategic assets without hedging language about continued U.S. partnership.
Why This Matters:
Export controls are a legitimate tool for protecting national security interests, but their implementation raises questions about unintended consequences and long-term strategic outcomes. Within two weeks of the ban, at least two companies—one in Tokyo and one in Beijing—have stepped into the market gap left by restricted U.S. access. Even if American companies could win back trust should the ban ever end, local alternatives trained to better understand local language and nuance are already filling the gap and building customer relationships. This dynamic illustrates a broader tension in technology policy: restrictions designed to preserve American competitive advantage may instead accelerate the development of competing ecosystems, ultimately fragmenting global AI development and reducing American influence over how these powerful tools are governed internationally. The question facing policymakers is whether export controls achieve their security objectives without triggering the technological decoupling and reduced global cooperation they aim to prevent.