Asian AI startups launch Mythos-like models as Anthropic’s export ban drags on
TL;DR
China’s 360 reportedly unveiled Tulongfeng, a security AI tool meant to find software vulnerabilities and positioned against Anthropic’s export-restricted Mythos. Tokyo-based Sakana AI launched Fugu, an agent-orchestration model that it claims can stand next to Fable 5 and Mythos Preview. The backdrop is a US order that has blocked Anthropic from giving global access to Mythos and Fable 5 for roughly two weeks.
Nauti's Take
The key question is not whether Fugu or Tulongfeng already match Mythos in practice. A lot of this is launch PR and geopolitical signaling.
But in enterprise buying, availability often beats peak capability: a slightly weaker model that will still be usable tomorrow can be more valuable than a stronger one exposed to policy shutdown risk. That is the dangerous part for US labs.
Briefingshow
Export controls are not just policy; they change buying behavior. If Asian enterprises and governments cannot rely on access to top US models, they will test local alternatives sooner and with more urgency. Even if Anthropic regains access later, trust may be harder to recover than benchmark gaps.
Sources