China's cybersecurity firm 360 Security Technology has developed its own AI security tool to counter Anthropic's Mythos, Reuters reported on June 24 local time.

360 founder Zhou Hongyi (저우 훙이) unveiled two AI security tools called "Yitian Tulong" at the ISC.AI 2026 cybersecurity conference in Beijing.

Yitian Tulong is a name taken from Chinese wuxia novels, meaning "the Yitian Sword and the Tulong Saber". "Tulongfeng", which automatically finds software vulnerabilities, was introduced as a "China version of Mythos". Yitianzhen automates cyber defence and incident response.

Zhou said, "The other side should not be the only one to have such powerful weapons that can change the landscape of cyber attack and defence." He warned that while U.S. companies can scan software and core systems with models like Mythos, China could face the risk of "one-way transparency" if it lacks comparable capabilities.

According to 360, Tulongfeng found 3,432 software vulnerabilities, 105 of which were confirmed by Chinese authorities. Reuters said it could not independently verify the claim.

Zhou acknowledged that Chinese models lag the United States by 20 to 30 percent in basic capabilities, but said, "We do not have the luxury of waiting until capabilities fully catch up." 360 said it chose an agent approach that combines security expertise, a vulnerability database and automation tools with models, instead of relying on top-performing models and computing resources as the United States does.