The global artificial intelligence landscape is witnessing a significant pivot, driven by recent US policy decisions. A critical move by the Trump Administration, specifically banning the export of Anthropic’s powerful Mythos AI model and its restricted version, Fable 5, to non-American entities, has sent ripples across the industry. This restriction, now two weeks in effect, has created a vacuum, and Asian AI startups are rapidly stepping up to fill the void with their own innovative solutions.
Asian Innovation Fills the Void
In a rapid response to the export ban, two prominent Asian AI firms have unveiled impressive new models. This past Wednesday, Chinese cybersecurity giant 360 reportedly launched Tulongfeng, an advanced AI tool designed to detect software vulnerabilities, asserting its capability to challenge Anthropic’s Mythos directly. This bold move signals a clear intent to establish domestic leadership in critical AI applications.
Earlier in the week, Tokyo-based Sakana AI introduced Fugu, a frontier AI model named after the Japanese blowfish. Sakana AI claims Fugu stands “shoulder-to-shoulder with leading models like Anthropic’s Fable 5 and Mythos Preview,” emphasizing its robust capabilities. Notably, Fugu is also engineered as an orchestration model, capable of managing and coordinating access to other AI systems via their APIs.
While a Sakana AI spokesperson told TechCrunch that Fugu’s launch was “entirely coincidental” to the US ban, the company has not shied away from capitalizing on the opportune timing. Their website proudly advertises “delivering frontier capability without the risk of export controls,” highlighting a key differentiator for businesses in the region. The underlying research for Fugu has been in development since last year, with findings presented at ICLR this spring, showcasing a foundational approach to delivering high-value AI.
Founded in 2023 by former Google researchers Ren Ito, Llion Jones, and David Ha, Sakana AI specializes in creating affordable generative AI models. These models are not only efficient with smaller datasets but are also meticulously optimized for the nuances of the Japanese language and culture. This localization offers a distinct advantage in regional markets.
Strategic Responses to Export Controls
Sakana AI is actively targeting Japanese businesses and government agencies that seek to mitigate their exposure to tightening international export controls. However, the company is careful not to declare a permanent shift away from American AI dominance in Asia. Co-founder Ren Ito articulated this perspective at the G7 summit in Evian, emphasizing that US models remain vital to Asia’s technological ecosystem.
“We’d characterize the current moment in those terms rather than as a permanent realignment toward any one set of players,” Sakana’s spokesperson reiterated. Ito further elaborated on this view in a Project Syndicate op-ed, urging the US government to prioritize preserving access for its closest allies. He argued that “AI should not become a technology that is hoarded; it should be one that is developed together.”
David Ha, co-founder and CEO of Sakana, sees Fugu as more than just a tactical move during a competitor’s vulnerability. He highlighted on X that “Orchestration Models are the next frontier, beyond bigger models,” advocating for a distributed approach to AI. Ha argued that relying on a single provider for national infrastructure presents an unacceptable risk, a point starkly underscored by the recent export controls.
“Access to top models can disappear overnight,” Ha warned, underscoring the fragility of concentrated power in AI. He believes that “collective intelligence is the practical hedge against this concentration of power,” promoting a vision where multiple models work in concert. This strategy aims to ensure continued access and resilience in the face of geopolitical uncertainties.
While Sakana positioned Fugu as a crucial hedging strategy, China’s 360 has adopted a more direct competitive stance. Beyond Tulongfeng, the company also unveiled Yitianzhen, designed to automate cyber defense and incident response. This launch was accompanied by a clear message from 360’s founder, Zhou Hongyi.
According to Reuters, Zhou described vulnerability-finding AI as a critical national strategic asset for China. He also raised concerns about “one-way transparency,” a scenario where some nations could access advanced vulnerability-detection capabilities while others could not, potentially creating significant security imbalances. This highlights a clear national security dimension to the race for AI supremacy.
Reshaping the Global AI Landscape
Anthropic, a leading US AI lab, had been on an impressive growth trajectory, reporting a run-rate revenue of over $47 billion in May 2026. The exact proportion of this revenue derived from Asian enterprise customers remains undisclosed, yet the recent export ban undeniably creates significant challenges for their global expansion. This policy decision could inadvertently catalyze the growth of localized AI ecosystems.
In the mere weeks since the export order took effect, companies like Sakana AI and 360 have swiftly moved to fill the emerging gap. These local alternatives, often trained to better understand regional languages, cultural nuances, and specific regulatory environments, are gaining crucial traction. Even if US companies manage to regain trust should the ban eventually lift, the landscape of AI adoption in Asia is already undergoing a profound transformation, moving towards more diversified and regionally focused solutions.
Source: TechCrunch – AI
