Asian AI startups launch Mythos-like models as Anthropic’s export ban drags on
TL;DR
China’s cybersecurity firm 360 unveiled Tulongfeng, an AI tool for automated vulnerability discovery that is positioned as a rival to Anthropic’s Mythos. Tokyo-based Sakana AI launched Fugu, a model it says can match Fable 5 and Mythos Preview while orchestrating access to other models through APIs for agent workflows. The market opening comes from policy: the US government has blocked global access to Anthropic’s Mythos and Fable 5 for two weeks, including a more restricted version.
Nauti's Take
This is more than a PR window for Asian startups. If a top model can disappear overnight for political reasons, procurement becomes risk management.
Sakana is selling the hedge elegantly; 360 is making the national-security pitch more bluntly. The hard point: buyers of critical AI systems are not only buying model quality, but access stability, language fit, cultural nuance, and political predictability.
Briefingshow
Export controls are meant to protect frontier models, but they can also push demand toward local alternatives. That is dangerous for US labs because trust and integrations can move quickly when companies and governments lose reliable access. Asia is not just getting substitute models here; it is gaining a stronger case for domestic AI infrastructure.
Sources