China launches Tulong Feng AI as Anthropic restricts Mythos
In brief
- Qihoo 360 founder Zhou Hongyi unveiled Tulong Feng at ISC.AI 2026 in Beijing on June 24 as China's Mythos alternative.
- Z.ai released GLM-5.2 under MIT license with no geographic restrictions or modification limits.
- GLM-5.2 costs $0.17 per vulnerability finding versus over $1 for Claude-based workflows.
Anthropic's Export Restrictions Spark Domestic Response
Anthropic's cybersecurity AI model Claude Mythos operates behind strict guardrails. The U.S. restricted access through a vetted coalition that includes Microsoft, Apple, and other major tech giants—but Chinese companies are excluded from Glasswing, Anthropic's partner program. Zhou characterized Mythos as cyber nuclear weapons in the AI age: an autonomous system that can find vulnerabilities, analyze them, and build attack chains without human direction.
That asymmetry prompted action. Qihoo 360 unveiled Tulong Feng, positioning it as a domestic alternative. According to Zhou's claims, the system has found a cumulative 3,432 vulnerabilities, with 105 confirmed by Chinese regulatory bodies. The company also rolled out Yitian Zhen, an automated defense platform, and launched Panshi Zhidun—a new domestic security coalition it calls the Shield of Bedrock.
Open-Weight Model Undercuts Proprietary Pricing
Z.ai, also known as Zhipu AI, took a different approach. The lab released GLM-5.2 under an MIT license with no subscription gates, no geographic restrictions, and freely modifiable by anyone. The move came shortly after the U.S. government pulled Mythos and Fable offline.
Performance benchmarks suggest the open model is competitive. In Semgrep's evaluation of insecure direct object reference detection, GLM-5.2 scored 39%, ahead of Claude Code in the same test. A separate Graphistry evaluation found it matched Claude Opus 4.8 on a capture-the-flag challenge.
Cost matters too. The price per finding for GLM-5.2 is around $0.17, versus over $1 for Claude-based workflows.
Z.ai co-founder Tang Jie called Anthropic's withdrawal of Mythos and Fable deeply regrettable. The open release signals Beijing's intent to prevent future restrictions from blocking access to critical security tools—and to prove that capability doesn't require a U.S. vendor.
