### Summary of Developments In response to a U.S. government export ban (enacted in June 2026) that restricts access to Anthropic’s advanced AI models—specifically **Mythos** and **Fable 5**—to non-U.S. entities, Asian AI startups are rapidly launching competitive alternatives. * **Sakana AI (Tokyo):** Launched **Fugu**, a frontier AI model designed for agent orchestration. It is marketed as a way for Japanese businesses to maintain "frontier-level" capabilities while hedging against the risk of sudden U.S. export controls. * **360 (Beijing):** Unveiled **Tulongfeng** (for vulnerability discovery) and **Yitianzhen** (for cyber defense). The company framed these as "national strategic assets," explicitly citing the risk of "one-way transparency" where advanced AI capabilities are restricted to certain global actors. ### Why It Matters * **Erosion of U.S. AI Dominance:** The export ban has created a vacuum in Asian markets. Local startups are filling this gap with models optimized for local languages and cultural nuances, potentially leading to a long-term shift away from reliance on U.S. AI providers. * **Strategic Autonomy:** The move highlights a growing trend of "AI sovereignty." Companies and governments in Asia are increasingly viewing reliance on a single, U.S.-based provider for critical infrastructure as a significant national security risk. * **Shift to Orchestration:** Rather than just trying to build larger models, companies like Sakana are focusing on "orchestration models"—systems that can coordinate multiple AI agents—as a way to build collective intelligence that is more resilient to individual model bans. ### Sources * **TechCrunch:** "Asian AI startups launch Mythos-like models as Anthropic's export ban drags on" (June 27, 2026). This report provides the primary account of the product launches and the strategic rationale from Sakana AI and 360. * **Reuters:** Cited by TechCrunch regarding 360 founder Zhou Hongyi’s comments on "one-way transparency." * **Project Syndicate:** Referenced for an op-ed by Sakana AI co-founder Ren Ito regarding the necessity of preserving AI access for U.S. allies. * **X (Twitter):** David Ha (CEO of Sakana AI) provided public commentary on the strategic importance of orchestration models as a hedge against concentrated power.