360 Security Unveils a Budget Chinese AI Hacker to Rival Anthropic’s Banned Tool
This is peak "we have Anthropic at home" energy. A Chinese tech giant just built its own AI cyber-weapon because they are terrified the US will be the only ones allowed to hack everyone else in the sandbox.
The cybersecurity firm 360 Security Technology officially launched a dual-agent AI system named Yintian Tulung. This dramatic naming choice is a direct nod to a classic Chinese wuxia novel, because nothing screams modern digital warfare quite like ancient mythical swords. Under this poetic umbrella, the digital toolkit splits tasks between two virtual soldiers: Tulungfeng hunts down system vulnerabilities, while Yitianzhen automates security defense and patch management.
During the ISC.AI 2026 conference in Beijing, the company's founder, Zhou Hongyi, confidently pitched his new creation as China’s direct answer to Anthropic's famous hacking AI, Mythos. However, instead of pretending his tech is a flawless alien artifact, Zhou took a refreshingly honest route. He openly admitted that China's baseline AI models lag behind American counterparts by about 20 to 30 percent, meaning his new cyber-warrior is essentially a cobbled-together "budget" edition.
To compensate for using a slightly dumber brain, 360 Security relies on clever engineering. Instead of training one super-intelligent AI hacker, they built a massive, automated assembly line. They hooked their model to a giant database of known software flaws, mixed in decades of real-world human hacking expertise, and set it to run on autopilot. While the Americans are busy raising a single genius hacker model, the Chinese are basically deploying a tireless, cheap robot army that works 24/7 without needing coffee breaks or complaining about overtime.
According to the company, Tulungfeng has already sniffed out 3,432 software vulnerabilities, out of which Chinese government authorities have officially confirmed 105. Independent watchdogs like Reuters haven't been able to verify these numbers, which is typical for cybersecurity claims coming out of Beijing. The real test will come when the low-hanging digital fruit is gone, forcing the AI to hunt for genuinely complex, hidden bugs.
The geopolitical panic behind this tool is incredibly real. Anthropic’s original Mythos was so terrifyingly good at finding zero-day exploits in major operating systems that US regulators panicked and banned it alongside its sibling model Fable 5. Zhou, who also sits on China's top political advisory body, basically made it clear that Beijing cannot afford a world of unilateral digital transparency, where only Washington has the master keys to everyone's servers.
Source: Reuters
Comments
This is where the magic happens: AI reads your discussion and rewrites the article based on the most interesting comments. Each strong comment adds points to the meter below. Once the meter is full, the article updates live — no page reload needed.