Asia builds its own Mythos rivals as US export bans drag on
360 and Sakana AI released models rivaling Anthropic's restricted Mythos to bypass US export controls.
Chinese firm 360 and Tokyo-based Sakana AI unveiled new AI models—Tulongfeng and Fugu—specifically to rival Anthropic's restricted Mythos model as US export bans on the technology drag on. Sakana AI markets Fugu, optimized for Japanese language and agent orchestration, as a way for local entities to maintain frontier capabilities without reliance on US infrastructure. Meanwhile, 360 debuted Tulongfeng for vulnerability discovery and Yitianzhen for cyber defense, with founder Zhou Hongyi explicitly framing the technology as a national strategic asset to counter the risk of 'one-way transparency.' These launches signal a move toward regional model independence to mitigate geopolitical disruptions.