Chinese AI models restrict network attack capabilities
The story so far
Alibaba, Tencent, and Moonshot AI among others in China are restricting network attack capabilities and strengthening export controls. Alibaba's Qwen4-Max filters network attack materials from training data and does not publicly disclose weights; Tencent's Hy4 removes exploit-generated data; Kimi K3.1 Pro achieves an internal vulnerability exploitation benchmark score of 81.0%.
AI-written from the reports · Updated 56 minutes ago
Coverage timeline
Follow the coverage to see every side of the story.
- ITHome · AIChinese AI models such as Alibaba Qwen4-Max, Tencent Hy4, and Moonshot AI Kimi K3.1 Pro restrict network attack capabilities and strengthen export controls
Chinese AI models such as Alibaba's Qwen4-Max, Tencent's Hy4, and Moonshot AI's Kimi K3.1 Pro are restricting network attack capabilities and strengthening export controls. Alibaba's Qwen4-Max filters network attack materials from training data and does not publicly disclose weights; Tencent's Hy4 removes exploit-generated data; Kimi K3.1 Pro scored 81.0% on internal exploit benchmark tests.
Heat over time
Not enough data yet to draw a trend.