Asian AI startups launch Mythos-like models (techcrunch.com)

Recently, two Asian startups launched AI models that aim to rival the capabilities of Anthropic’s advanced models, Mythos and Fable. Chinese cybersecurity firm 360 introduced Tulongfeng, an AI tool designed for cybersecurity that can reportedly compete effectively with Mythos. Meanwhile, Tokyo-based Sakana AI revealed Fugu, a model tailored for Japanese language and culture, optimized for use with small datasets. The launch of these models coincides with the U.S. government's export ban on Mythos and Fable, which has shifted focus to local alternatives that address regional needs and reduce reliance on U.S. technology. The significance of these developments extends beyond mere competition; they reflect a changing landscape in the global AI arena prompted by geopolitical tensions. Sakana’s Fugu is positioned as a flexible solution that orchestrates access to various AI models, acting as a hedge against the unpredictability of U.S. export controls. 360's Tulongfeng, on the other hand, emphasizes the idea of vulnerability detection as a national strategic asset. These launches highlight a potential shift in AI capabilities, where Asian firms are stepping in to fill voids left by U.S. restrictions, suggesting a trend toward localized AI solutions that not only mitigate geopolitical risks but also cater to specific regional demands.