The landscape of global artificial intelligence is undergoing a profound structural realignment, catalyzed by sweeping regulatory actions from Washington and rapid technological counter-mobilization across Asia. Just weeks after the U.S. federal administration enacted stringent export controls barring non-Americans from accessing Anthropic’s powerhouse cybersecurity models—namely Mythos Preview and its restricted counterpart, Fable 5—major technology firms in Tokyo and Beijing have rolled out indigenous alternatives.
This swift emergence of regional competitors underscores the growing fragility of centralized technological dependencies. As international supply chains fracture along geopolitical fault lines, the race for artificial intelligence sovereignty has shifted from a theoretical debate into an urgent operational necessity for foreign governments and enterprises alike.
The Catalyst: The U.S. Ban on Anthropic’s Frontier Models
The current wave of disruption traces back to an executive decision made by the U.S. administration, which placed strict limitations on the global distribution of Anthropic’s most advanced models. The restrictions specifically targeted Anthropic’s Mythos Preview and Fable 5, models recognized industry-wide for their unprecedented capabilities in autonomous vulnerability discovery, complex agentic reasoning, and sophisticated cyber operations.
While Anthropic was riding an unprecedented wave of commercial momentum—having reported that its run-rate revenue crossed a staggering $47 billion in May 2026 alongside a looming near-trillion-dollar valuation—the geopolitical friction point arrived abruptly. The federal decree blocked international deployment of these systems outside of U.S. borders, citing national security implications and the potential dual-use nature of advanced offensive and defensive cyber intelligence models.
For international allies and strategic competitors alike, the sudden cutoff served as an alarming wake-up call. Relying on a single domestic market for critical national infrastructure intelligence suddenly presented an unacceptable existential risk. Within days of the restriction taking effect, the vacuum left by Anthropic’s withdrawal began to attract aggressive, localized counter-deployments designed to insulate foreign economies from future supply shocks.
Sakana AI Introduces Fugu: Tokyo’s Answer to Orchestrated Intelligence
In Tokyo, Japanese artificial intelligence pioneer Sakana AI capitalized on the shifting tides with the mid-week release of Fugu, a frontier AI model named after the notoriously difficult-to-prepare Japanese blowfish. Founded in 2023 by Google alumni David Ha and Llion Jones, alongside former Mercari and Stability AI executive Ren Ito, Sakana has carved out a distinct market niche by building affordable, resource-efficient generative AI models optimized for smaller datasets, regional linguistic nuances, and Japanese cultural contexts.
According to Sakana AI, the timing of Fugu’s commercial release was entirely coincidental. Company representatives noted that the foundational research behind the model had been presented earlier this spring at the International Conference on Learning Representations (ICLR), and that development had been underway since the previous year. Nevertheless, the company found itself in an advantageous position to market Fugu as an immediate alternative for Japanese corporations and government agencies seeking to minimize their exposure to unpredictable regulatory shifts.
Sakana’s leadership emphasized that Fugu is designed not merely as a localized replica of U.S. tech, but as a technological leap forward focusing on agent orchestration. Unlike monolithic systems designed to handle every computational task independently, Fugu acts as a supervisory framework capable of seamlessly orchestrating access to a multitude of other models via their respective application programming interfaces (APIs).
In a series of public statements, CEO David Ha underscored that "orchestration models are the next frontier, beyond bigger models." Ha argued on social media that the vulnerabilities exposed by sudden export controls have made reliance on a single provider untenable. "Access to top models can disappear overnight," Ha wrote. "Collective intelligence is the practical hedge against this concentration of power."
Despite the commercial tailwinds driving interest toward Fugu, Sakana’s leadership has maintained a measured diplomatic stance. During the G7 summit in Evian, France, co-founder Ren Ito addressed global leaders, asserting that U.S. models remain vital to the technological ecosystem of Asia and framing the current period as a temporary adaptation rather than a permanent, irrevocable decoupling.
In a subsequent op-ed published via Project Syndicate, Ito urged Western policymakers to prioritize the preservation of access for trusted international allies, arguing that artificial intelligence should function as a collaborative global asset rather than a hoarded proprietary weapon.
Beijing Responds: 360 Unveils Tulongfeng and Yitianzhen
While Tokyo’s approach leaned toward diversification and cooperative orchestration models as a hedge against volatility, Beijing’s response reflected a more direct and uncompromised pursuit of technological self-reliance. On Wednesday, major Chinese cybersecurity firm 360 formally unveiled Tulongfeng, an advanced AI model engineered to compete directly with Anthropic’s restricted Mythos architecture.
Alongside Tulongfeng, 360 introduced Yitianzhen, a companion model specifically designed to automate comprehensive cyber defense protocols and real-time incident response. Together, the twin tools represent a significant leap in China’s domestic defensive and offensive algorithmic capabilities.
The launch carried unmistakable strategic undertones. According to reports from Reuters, 360 founder Zhou Hongyi framed automated vulnerability-discovery systems as foundational national security assets. Hongyi explicitly warned against the dangers of "one-way transparency"—a scenario in which adversarial nations maintain absolute dominance over advanced vulnerability-detection technologies while leaving other global actors blind and defenseless.
Unlike Sakana AI, which welcomed international collaboration while building protective local alternatives, 360’s launch signals a hardening of technological autonomy within mainland China. Industry analysts note that Chinese firms have systematically accelerated domestic substitution strategies in response to successive rounds of U.S. semiconductor sanctions and software export blocks. The arrival of Tulongfeng confirms that this substitution trend has officially expanded into the domain of frontier agentic reasoning and automated cyber warfare.
Chronology of Events: A Rapidly Escalating Timeline
To understand the current state of international AI fragmentation, it is essential to trace the sequence of developments that compressed years of expected geopolitical realignment into a matter of weeks:
- May 2026: Anthropic announces historic financial milestones, reporting that its run-rate revenue has surpassed $47 billion as it prepares for a monumental initial public offering near a $1 trillion valuation.
- Early June 2026: The U.S. administration issues an emergency directive restricting the deployment of Anthropic’s advanced Mythos Preview and Fable 5 models outside the United States, citing national security concerns over cyber intelligence capabilities.
- Mid-June 2026: G7 leaders convene in Evian, France, where cross-border AI access, sovereign computing infrastructure, and export controls dominate high-level diplomatic discussions. Sakana co-founder Ren Ito publicly advocates for open technological sharing among allied nations.
- Late June 2026: Sakana AI officially launches Fugu, promoting its agent-orchestration capabilities as an operational hedge against sudden regulatory cutoffs.
- Wednesday, June 24, 2026: Chinese cybersecurity giant 360 unveils Tulongfeng and Yitianzhen, explicitly positioning the models to rival Anthropic’s restricted technologies and warning against asymmetrical global transparency.
Economic and Strategic Implications for the Global Tech Market
The concurrent rise of Sakana’s Fugu and 360’s Tulongfeng marks a critical turning point in the commercial lifecycle of generative artificial intelligence. For the past several years, the global enterprise market operated under the assumption of a centralized technological hierarchy anchored by a handful of elite American laboratories.
The enforcement of the Anthropic export ban has effectively shattered that paradigm. Even if political tensions were to de-escalate and export controls were eventually lifted, regional markets have already absorbed a vital lesson in vulnerability. International enterprises and government ministries are unlikely to reverse investments in localized sovereign AI solutions that offer superior comprehension of regional languages, cultural nuances, and domestic regulatory frameworks.
Furthermore, the pivot toward orchestration-based models—pioneered by startups like Sakana—heralds a structural shift in how businesses utilize artificial intelligence. Rather than depending on a single monolithic model to handle complex automated tasks, future corporate infrastructures will likely rely on federated networks of specialized, multi-vendor agents capable of routing queries across diverse domestic and international APIs.
Ultimately, the actions taken by Tokyo and Beijing this week illustrate that artificial intelligence has officially completed its transition from a commercial consumer novelty into foundational geopolitical infrastructure. As local alternatives mature to fill the voids left by Western regulatory restrictions, the vision of a unified global internet and a single shared AI ecosystem is rapidly giving way to a fractured, multi-polar technological order.
