The global artificial intelligence landscape is undergoing a profound structural realignment following sweeping U.S. export restrictions on cutting-edge American AI models. In the wake of the Trump administration’s decision to restrict non-American access to Anthropic’s powerhouse cybersecurity models, Mythos and Fable 5, international competitors have moved swiftly to fill the vacuum. Within a single week, Tokyo-based startup Sakana AI and Chinese cybersecurity giant 360 unveiled advanced frontier models designed to match or exceed the capabilities of the banned U.S. technology. This rapid development highlights a growing trend toward technological sovereignty, as Asian markets seek to mitigate the operational risks posed by sudden shifts in American trade policy.
The Catalyst: The U.S. Export Ban on Anthropic’s Mythos and Fable 5
The geopolitical friction driving these regional developments stems from a policy decision enacted by the United States federal government. Approximately two weeks prior to the recent Asian product launches, the Trump administration issued an emergency access order restricting Anthropic—one of the world’s leading artificial intelligence laboratories—from distributing its most advanced cybersecurity-focused AI models globally. Specifically, the regulatory block targeted Mythos and its more restricted counterpart, Fable 5.
The U.S. government justified the restrictions by pointing to the extraordinary power and dual-use potential of these models, which possess unprecedented capabilities in identifying software vulnerabilities and automating complex cyber operations. Washington argued that maintaining stringent control over such high-level assets is vital for national security. However, the extraterritorial nature of the ban sent shockwaves through international markets. For allies and strategic competitors alike, the sudden denial of access laid bare the inherent vulnerabilities of relying on a single nation’s technological infrastructure for critical digital operations.
Anthropic, meanwhile, had been riding an unprecedented wave of commercial momentum. Prior to the export restrictions, the company reported that its annualized run-rate revenue had crossed an astonishing $47 billion in May 2026, pushing its valuation near the $1 trillion mark ahead of a highly anticipated initial public offering. While the exact proportion of revenue derived from Asian enterprise customers remains proprietary, the sudden policy shift threatens to disrupt established supply chains and force international clients to reconsider their long-term software dependencies.
Sakana AI Introduces Fugu: Tokyo’s Answer to Frontier Restrictions
Stepping into this volatile environment, Tokyo-based Sakana AI officially launched its latest frontier model, Fugu—named after the Japanese word for blowfish—earlier in the week. Founded in 2023 by Google alumni David Ha and Llion Jones alongside former Mercari and Stability AI executive Ren Ito, Sakana AI has carved out a distinct niche by developing cost-effective generative AI models optimized for smaller datasets, local languages, and specific cultural nuances.
According to company leadership, Fugu stands shoulder-to-shoulder with industry-leading benchmarks, matching the performance profiles of Anthropic’s Fable 5 and Mythos Preview. Unlike traditional monolithic models, Fugu is specifically engineered for multi-agent ecosystems, possessing the native ability to orchestrate access to various other models through their respective application programming interfaces (APIs).
While market observers immediately drew parallels between the timing of Fugu’s release and the ongoing U.S. export ban, Sakana AI representatives insisted that the debut was entirely coincidental. A company spokesperson noted that the underlying research for Fugu had been presented at the International Conference on Learning Representations (ICLR) earlier in the spring, representing the culmination of development efforts that began the previous year.
Nevertheless, Sakana has not hesitated to leverage the geopolitical moment. The startup’s official marketing materials now explicitly promote the platform as a means of "delivering frontier capability without the risk of export controls." By positioning Fugu as a strategic hedge rather than a wholesale replacement for American technology, Sakana aims to reassure Japanese businesses and government agencies that are increasingly anxious about tightening foreign trade barriers.
Strategic Perspectives from Sakana’s Leadership
The launch of Fugu also served as a backdrop for high-level diplomatic and economic commentary from Sakana’s founders. Ren Ito addressed global leaders at the G7 summit in Evian, where discussions regarding artificial intelligence access and export controls dominated the agenda. Ito advocated for a collaborative, open approach to technological advancement, arguing that digital sovereignty should be built on a foundation of diverse options rather than restrictive ownership.
In a subsequent op-ed published by Project Syndicate, Ito urged the U.S. federal government to recognize that preserving technological access for its closest allies must remain a primary policy objective. He cautioned against treating advanced AI as a commodity to be hoarded, asserting instead that sustainable innovation depends on joint development.
Expanding on these themes, CEO David Ha emphasized that the proliferation of orchestration models represents the next logical frontier in artificial intelligence development, surpassing the race for sheer parameter size. Writing on social media platform X, Ha argued that depending on a single foreign provider for national infrastructure introduces an unacceptable level of operational risk.
"Access to top models can disappear overnight," Ha wrote. "Collective intelligence is the practical hedge against this concentration of power." Through this lens, Fugu is designed not to isolate Asian markets from Western innovation, but to create a resilient, decentralized framework where organizations can fluidly switch between multiple AI providers to insulate themselves from geopolitical turbulence.
China’s 360 Unveils Tulongfeng and Yitianzhen: A Direct Challenge
While Tokyo-based Sakana framed its new release as a risk-mitigation strategy to preserve access, Chinese cybersecurity titan 360 adopted a much more assertive posture. On Wednesday, the firm officially unveiled Tulongfeng, an advanced AI tool engineered to compete directly with Anthropic’s Mythos. Alongside Tulongfeng, 360 also introduced Yitianzhen, a companion system designed to automate cyber defense protocols and real-time incident response.
The rollout of these tools was accompanied by strong rhetoric from 360 founder Zhou Hongyi. According to reports from Reuters, Zhou characterized vulnerability-finding artificial intelligence as an essential national strategic asset. He heavily criticized what he termed the risk of "one-way transparency"—a scenario where select global powers retain exclusive access to advanced offensive and defensive cyber capabilities while leaving other nations vulnerable.
Unlike Sakana, which continues to emphasize the ongoing importance of U.S. models within Asian markets, 360’s aggressive product rollout signals an accelerated drive toward domestic self-reliance. By pairing offensive vulnerability discovery with automated defense systems, the Chinese firm is signaling that domestic enterprises and state entities will no longer wait for foreign policy shifts to secure their digital borders.
Implications and the Future of Global AI Governance
The simultaneous emergence of Sakana’s Fugu in Japan and 360’s security suite in China underscores the unintended consequences of unilateral trade restrictions in the high-tech sector. While export controls are ostensibly designed to maintain strategic technological superiority and safeguard national security, they concurrently stimulate fierce competitive pressures abroad.
Local alternative models, specifically trained to navigate regional languages, regulatory frameworks, and cultural nuances, are rapidly filling the vacuum left by restricted American platforms. Even if current diplomatic tensions were to ease and export bans were eventually lifted, international customers who experienced operational disruptions may remain hesitant to rebuild their core infrastructure around single-source American providers.
Furthermore, the pivot toward agentic orchestration—championed by Sakana AI—suggests a structural shift in how enterprises will consume artificial intelligence moving forward. Rather than depending on a monolithic proprietary model from a single vendor, organizations are increasingly incentivized to build modular, multi-model architectures that can dynamically route tasks across diverse geographic and corporate ecosystems.
As this fragmented landscape solidifies, the global artificial intelligence market is transitioning away from a centralized, U.S.-dominated paradigm toward a multi-polar framework. In this new era, technological capability is no longer measured solely by computational power or raw parameter counts, but by resilience, adaptability, and the capacity to withstand the shifting winds of geopolitical strategy.
