Sakana AI and 360 launch models as US export bans persist

On Wednesday, 360, a cybersecurity firm in China, reportedly unveiled Tulongfeng, an AI tool it says can compete with Anthropic’s Mythos. This launch follows the release of Fugu by Japan-based startup Sakana AI, which was founded in 2023. Sakana AI asserts that Fugu performs similarly to Anthropic’s Fable 5 and Mythos Preview. These new products from Asian companies arrive as the US government has banned Anthropic from providing global access to its Mythos and Fable models.

The US export ban, which took effect two weeks ago, restricts access to Anthropic’s Mythos, the cybersecurity-focused AI model that is reportedly so powerful, the Trump Administration has currently banned it and its more restricted version, Fable 5, from the hands of non-Americans. In response to these restrictions, 360 founder Zhou Hongyi flagged what he called the risk of “one-way transparency”, a situation in which some actors could access advanced vulnerability-detection capabilities while others could not.

Sakana AI co-founder and CEO David Ha described Fugu as an Orchestration Model—an AI model designed to coordinate agent usage among many models—which he calls the next frontier beyond bigger models. Ha warned that access to top models can disappear overnight, arguing that “Collective intelligence is the practical hedge against this concentration of power.” A spokesperson for Sakana AI, co-founded by Ren Ito, stated that the timing of the Fugu release was coincidental, noting that the company had been building the model since last year and presented the underlying research this spring. While Sakana AI is targeting Fugu at Japanese businesses and government agencies, the spokesperson noted that U.S. models remain important to Asia, characterizing the current situation as a temporary moment rather than a permanent realignment.

The emergence of local alternatives comes during a period of growth for US providers. Anthropic reported that its run-rate revenue reached $47 billion in May 2026. However, as export controls persist, local alternatives in China and Japan are stepping in to serve enterprise customers in Asia who seek to reduce their exposure to tightening US trade policies.

Why it matters

As US export controls restrict access to American AI models, local alternatives are emerging in Asia. This shift could reduce the long-term reliance of enterprise customers in Asia on US technology providers.