Asian AI Startups Unveil Alternative Models Amid Anthropic Export Ban
모델 출시 | Sun Jun 28 2026 00:00:00 GMT+0000 (Coordinated Universal Time) | 2 sources
China's 360 and Japan's Sakana AI released frontier-grade alternative models against the backdrop of the Anthropic Mythos and Fable 5 export ban.
Analysis
[360] unveiled Tulongfeng cybersecurity AI model [1][2]
- Positioned as a direct competitor to Anthropic Mythos
- Developed by a Chinese cybersecurity company
- Announced amid the U.S. export ban on Mythos and Fable 5
[Sakana AI] launched frontier model Fugu [1][2]
- Claims to stand shoulder-to-shoulder with Anthropic Fable 5 and Mythos Preview
- Designed for agents to orchestrate other model APIs
- Small-dataset-friendly model optimized for Japanese language and culture
[Sakana AI] marketed frontier performance free of export control risk [1][2]
- Website advertises "frontier capabilities with no export control risk"
- Primarily targets Japanese enterprises and government agencies
- Explained the launch timing as "entirely coincidental"
[David Ha (Sakana AI CEO)] positioned orchestration models as the next frontier [1][2]
- Fugu is designed to coordinate agentic use across multiple models
- Stated "orchestration models are the next frontier beyond larger models"
- Emphasized significance beyond simply filling the gap left by U.S. competitors
[Ren Ito (Sakana AI co-founder)] called for guaranteed AI access through G7 Summit and Project Syndicate op-ed [1]
- Discussed AI access and export controls at the Evian G7 Summit
- Presented preserving access for U.S. allies as a top priority
- Argued "AI is not a technology to be stockpiled but to be developed together"