360 Security Technology unveiled Yitian Tulong at ISC.AI 2026 in Beijing — an AI-powered vulnerability discovery and defense platform framed as China’s answer to Anthropic’s Mythos.

Founder Zhou Hongyi described “one-way transparency” risk: if only one country possesses AI capable of scanning global digital infrastructure, that country gains strategic advantage.

Yitian Tulong comprises:

  • Tulongfeng: automated vulnerability discovery (“China’s version of Mythos”) — claims 3,432 vulnerabilities found, 105 government-confirmed (unverified by Reuters)
  • Yitianzhen: automated cyber defense and incident response

360 acknowledged Chinese foundation models lag US rivals by 20-30% in raw computing power due to chip export restrictions. Strategy: combine AI models with cybersecurity expertise, vulnerability databases, and automated tooling rather than competing on model size alone.

Trigger: US June 2026 export controls suspending foreign access to Anthropic Mythos and Fable 5.

Zhou: “This kind of powerful weapon that can change the landscape of cyber offence and defence cannot be held only by others.”