Asian AI startups launch Mythos-like models
Points and comments are a snapshot, not live.
Asian AI startups launch models comparable to Anthropic's banned Mythos.
Chinese cybersecurity firm 360 unveiled Tulongfeng, an AI vulnerability-finding tool it says matches Anthropic's Mythos. Tokyo-based Sakana AI launched Fugu, an orchestration model for agents, claiming parity with Fable 5 and Mythos Preview. Both launches follow the U.S. export ban on Anthropic's top models. Sakana, founded by former Google researchers, targets Japanese businesses seeking sovereign AI. 360's founder called vulnerability-detection AI a national strategic asset, warning of "one-way transparency."
What commenters are saying
Commenters split on the significance of these launches. Some argued the U.S. ban backfired, accelerating Asian alternatives and threatening American AI leadership. Others expressed skepticism, noting Sakana had previously retracted performance claims, and questioning whether models from unproven companies could truly match Mythos. Several flagged the broader risk: technology concentration under export controls can disappear overnight, making local hedges necessary. A minority defended Anthropic, blaming regulatory fearmongering for the restrictions.