China Challenges American Mythos with New AI Weapons

Chinese security company 360 Security Technology unveiled new AI tools at the recent ISC.AI 2026 conference in Beijing under the collective name “Yitian Tulong”. This historical name, taken from the famous Chinese martial arts novel, is provided by the “Tulongfeng” system. The advanced system automatically detects security holes in software. The “Yitianzhen” defense system was created to respond to attacks automatically.

The company's founder, Zhou Hongyi, stressed that such a powerful weapon, which changes the rules of the game in cyber defense and offense, should not be left to Western competitors. The Mythos model, announced by the American Anthropic in April, has caused unrest among experts, as it can detect thousands of critical vulnerabilities in operating systems and browsers with extreme ease, which could seriously strengthen the power of hacking attacks. The US has therefore already restricted the export of this technology to protect national security.

As China faces tough U.S. restrictions on the import of advanced chips, making it difficult for domestic companies to develop equally powerful AI models, they have had to take a different path. Zhou acknowledges that China’s basic models are still about 20 to 30 percent behind, but warns that the country can no longer sit idly by. Instead of relying solely on processing power and chips, 360 has combined AI with vast databases of vulnerabilities and expertise.

Although it is a partially experimental approach, the company claims that Tulongfeng has already successfully identified 3,432 software bugs, of which 105 have been officially confirmed by Chinese authorities. These figures remain unverified. Nevertheless, 360, which made its name developing antivirus software, now offers a comprehensive solution aimed at protecting state infrastructure.