Two Asian AI companies launched products this week that position themselves as alternatives to Anthropic's suspended Mythos and Fable 5 models. Tokyo-based Sakana AI released Fugu, an orchestration model it says matches Fable 5 on key benchmarks, while Beijing cybersecurity firm 360 Security unveiled Tulongfeng, a vulnerability-discovery tool it claims can rival Mythos. Both launches arrived as the US government's export ban on Anthropic's most capable models entered its third week with no resolution in sight.

Sakana's approach is unusual. Rather than training a new frontier model from scratch, it built a seven-billion-parameter orchestrator whose job is to decide which external model should handle each part of a problem. Fugu routes tasks across a pool of available models, assembling and coordinating them as a team, and the company says the result matches the performance of systems that cost orders of magnitude more to train.

Sakana was founded in 2023 by Llion Jones, a co-author of the original Transformer paper at Google, David Ha, a former Google Brain researcher, and Ren Ito, a former Japanese diplomat. The company raised $135 million in a Series B round in November 2025 at a valuation of nearly $3 billion, and its research on multi-model orchestration was presented at ICLR this spring.

A Sakana spokesperson told TechCrunch the timing was “entirely coincidental,” but the company has not been shy about capitalising on the moment. Its website advertises “delivering frontier capability without the risk of export controls.” Co-founder David Ha framed the product as more than opportunism, arguing that relying on a single provider for national infrastructure is a risk the export ban made impossible to ignore.

China's 360 Security was more direct. At the ISC AI 2026 cybersecurity conference in Beijing on Tuesday, founder Zhou Hongyi unveiled two tools grouped under the name Yitian Tulong. Tulongfeng targets automated vulnerability discovery, while Yitianzhen is designed to automate cyber defence and incident response.

Zhou described vulnerability-finding AI as a national strategic asset and warned of what he called “one-way transparency,” a situation in which some countries can examine software for weaknesses while others cannot.

Zhou conceded that Chinese models still lag behind American ones, estimating a gap of 20 to 30 percent in base capability, but argued that waiting for parity was not an option. The company says Tulongfeng has identified 3,432 software vulnerabilities so far, including 105 confirmed by Chinese authorities. Reuters said it could not independently verify those claims.

The two launches reflect a dynamic that the Anthropic ban has accelerated across Asia. India is debating a $5 billion sovereign AI fund, and AI access was a central topic at the G7 summit in Evian last week, where Sakana's Ren Ito was among the business leaders invited to participate.

Ito later wrote in an op-ed for Project Syndicate that the US government should prioritise preserving access for its closest allies. He argued that AI should be developed together, not hoarded, and urged Washington to consider the consequences of treating frontier models as export-controlled weapons rather than shared infrastructure.

Neither Sakana nor 360 is claiming to replace American frontier AI outright. Sakana frames Fugu as a hedge, a way to preserve capability even when access to a single provider disappears overnight. Ito told TechCrunch that “US models remain important to Asia” and characterised the current moment as something short of a permanent realignment.

But the commercial reality is moving faster than the diplomacy. Anthropic's run-rate revenue crossed $47 billion in May, and how much of that depends on Asian enterprise customers is not publicly known. What is known is that two weeks of export restrictions have already produced two concrete competitors, one in an allied capital and one in a rival's, both marketing themselves on a promise Washington's ban made for them: that dependence on American AI carries a risk no amount of capability can offset.