The global race for artificial intelligence supremacy has entered a precarious new phase. Following the Trump Administration’s recent decision to impose sweeping export controls on Anthropic’s flagship cybersecurity models—Mythos and its restricted counterpart, Fable 5—the international AI market is witnessing a rapid, localized pivot. As U.S. restrictions effectively wall off top-tier American technology from non-domestic entities, regional powerhouses in Tokyo and Beijing are accelerating their own sovereign AI initiatives.
This geopolitical friction has created an immediate vacuum, prompting Tokyo-based startup Sakana AI and Chinese cybersecurity giant 360 to unveil competitive alternatives. The move suggests a fundamental shift: the era of reliance on a singular, U.S.-led AI infrastructure may be drawing to a close, replaced by a fragmented landscape of regional "AI sovereignty."
A Chronology of the Shift
The current instability stems from a series of rapid-fire developments over the past month.
- Mid-June 2026: The Trump Administration formally issues an executive order restricting the global access of Anthropic’s Mythos and Fable 5 models. These tools, highly prized for their advanced cybersecurity and vulnerability-detection capabilities, are deemed critical national security assets, now off-limits to non-U.S. entities.
- June 17, 2026: During the G7 summit in Evian, discussions surrounding AI governance and export controls reach a fever pitch, highlighting the growing tension between national security and technological cooperation.
- June 22, 2026: Sakana AI, a rising star in the Tokyo tech scene, launches its frontier model, "Fugu," explicitly marketed as a high-performance alternative to Anthropic’s restricted offerings.
- June 24, 2026: Chinese cybersecurity leader 360 unveils "Tulongfeng" and "Yitianzhen," two sophisticated AI tools designed to challenge Anthropic’s dominance in the cybersecurity sector, framing them as essential strategic assets for the Chinese state.
The Rise of Sovereign AI Alternatives
The swift arrival of Fugu and Tulongfeng is not merely a reaction to the U.S. ban; it is the manifestation of a long-standing desire for technological independence.
Sakana AI’s "Fugu": The Agentic Hedge
Sakana AI, founded in 2023 by former Google researchers Ren Ito, Llion Jones, and David Ha, has positioned Fugu as more than just a copycat. Designed for "orchestration," Fugu is built to manage agentic workflows, allowing businesses to interface with various AI models via APIs.
Co-founder David Ha argues that the reliance on a single, U.S.-based provider for national infrastructure is a precarious strategy. "Access to top models can disappear overnight," Ha noted on social media, emphasizing that "collective intelligence is the practical hedge against this concentration of power." By focusing on small datasets and deep optimization for the Japanese language and cultural context, Sakana is carving out a niche that emphasizes reliability over global dominance.
360’s "Tulongfeng": Security as a Strategic Asset
In contrast to Sakana’s collaborative approach, the Chinese firm 360 has taken a more aggressive stance. Their new tools, Tulongfeng and Yitianzhen, are designed for automated software vulnerability discovery and cyber defense.
The rhetoric accompanying this launch underscores a deepening divide. 360 founder Zhou Hongyi highlighted the dangers of "one-way transparency"—a scenario where Western powers utilize AI to monitor or penetrate foreign networks while denying those same capabilities to others. For 360, the development of these tools is a matter of national defense, framed as a "strategic asset" that ensures China remains on equal footing in the digital arms race.
Supporting Data and Economic Context
The impact of these export controls is amplified by the sheer scale of the companies involved. Anthropic, which recently reported a staggering $47 billion run-rate revenue as of May 2026, has seen its valuation climb toward the $1 trillion mark. However, the loss of access to Asian markets—historically significant hubs for enterprise software—could disrupt this growth trajectory.
While the exact percentage of Anthropic’s revenue derived from Asian clients remains proprietary, the market reaction is telling. Startups like Sakana AI, which secured a $2.65 billion valuation in their Series B round late last year, are now aggressively targeting the enterprise segment that once relied on U.S. frontier models. These local alternatives, which are often more affordable and culturally localized, are finding a ready-made audience among businesses and government agencies desperate to avoid the disruption caused by shifting U.S. policy.
Official Responses and Diplomatic Friction
The diplomatic fallout from the U.S. ban has been significant. Ren Ito of Sakana AI has been a vocal proponent of a more open approach, arguing that the U.S. should prioritize maintaining access for its closest allies rather than "hoarding" AI capabilities.
In an op-ed published in Project Syndicate, Ito stated, "AI should not become a technology that is hoarded; it should be one that is developed together." His perspective, echoed at the G7 summit, is that the current moment should not be viewed as a permanent realignment of the global AI order, but as a warning. The message to Washington is clear: if the U.S. continues to utilize AI as a geopolitical cudgel, it risks forcing its allies and trading partners into the arms of local, sovereign AI ecosystems that are increasingly capable of filling the gap.
Implications for the Future of Global AI
The emergence of Fugu and Tulongfeng signals a fundamental change in the trajectory of the AI industry. We are witnessing the end of a "monoculture" of AI development, where a few Silicon Valley labs dictated the standards and capabilities for the entire world.
1. The Fragmentation of Standards
As regions develop their own models, we are likely to see a divergence in AI performance, safety standards, and ethical guidelines. A model optimized for the Japanese regulatory environment or the Chinese security apparatus will naturally prioritize different outcomes than one built in San Francisco.
2. The Rise of "Orchestration"
As David Ha pointed out, the future may not belong to the largest model, but to the best "orchestrator." Companies that can effectively manage a fragmented landscape—using different models for different tasks—will likely hold the advantage over those tethered to a single, potentially unreliable, provider.
3. The End of U.S. Technological Hegemony
The most profound implication is the erosion of the assumption that American AI is the only choice. By restricting access to its most advanced tools, the U.S. has inadvertently accelerated the maturity of its competitors. Even if the current ban were lifted tomorrow, the trust deficit created by the sudden cutoff of service will likely lead many international firms to permanently diversify their AI procurement.
4. Security as a Catalyst for Sovereignty
Finally, the "weaponization" of AI for cybersecurity—as seen in the 360 rollout—means that AI is no longer just a productivity tool; it is a critical component of national security. When security, defense, and intelligence functions are on the line, governments will almost always choose the model they can control over the model they must request permission to use.
As the dust settles on this latest round of export controls, it is clear that the AI landscape is becoming more complex, more regionalized, and significantly more competitive. The U.S. move may have been intended to slow the advancement of rivals, but in the short term, it has served as a powerful catalyst for global innovation, proving that in the digital age, isolation is often the most effective driver of technological self-reliance.
rifanmuazin
Nana
Nana Muazin
