China’s 360 says it has developed tools to match Anthropic’s Mythos – Tech Digest

Share

Chinese cybersecurity firm 360 Security Technology has announced the development of a new suite of AI-powered tools, which it claims match the capabilities of Anthropic’s powerful “Mythos” system.

The move signals an escalation in the digital arms race, with 360’s founder, Zhou Hongyi, characterising the technology as a necessary “strategic asset” for China’s national security.

Unveiled at the ISC.AI 2026 conference in Beijing, 360’s new AI tools are collectively branded as “Yitian Tulong” – a reference to a classic Chinese martial arts novel.

The primary component, “Tulongfeng,” is specifically designed to discover software vulnerabilities automatically, mirroring the functionality of Anthropic’s Mythos. A second tool, “Yitianzhen,” focuses on automating cyber defence and incident response.

The development comes amid heightening global concern regarding the potential for AI to supercharge cyberattacks. Anthropic’s Mythos was famously restricted by the US government due to national security fears after it demonstrated an unprecedented ability to identify thousands of flaws in critical software.

Zhou, who has frequently argued that China risks “one-way transparency” if it lacks equivalent technology while US entities possess it, described the emergence of such AI as the new frontier of strategic deterrence.

“This kind of powerful weapon that can change the landscape of cyber offence and defence cannot be held only by others,” Zhou stated, according to a transcript of his speech. He likened the capability to a “cyber nuclear weapon,” emphasising that China cannot afford to wait until its base models fully catch up to US standards before deploying such defences.

Strategic differences

While acknowledging that domestic AI models still lag behind their American counterparts by roughly 20% to 30% in raw base capability, Zhou outlined a different strategy for closing the gap. Instead of attempting to replicate the US approach – which relies on the most advanced chips and massive computing power – 360 is focusing on an “agent” route.

By combining AI models with specialised security expertise and vulnerability databases, the company claims it has created a “complete machine” that can operate 24 hours a day. “If the US route is to cultivate a genius hacker, 360’s route is to organise a professional attack-and-defence team,” Zhou explained.

360 claims that Tulongfeng has already identified over 3,400 vulnerabilities. While these claims remain unverified, the announcement arrives as cybersecurity leaders from the “Five Eyes” nations warn that adversaries could be mere months away from launching AI-powered attacks that threaten global critical infrastructure.

For latest tech stories go to TechDigest.tv

Related Posts