Tencent launches Hy4 open-source AI model focused on defensive cybersecurity
Technology
Tencent just launched Hy4, an open-source AI model built with cybersecurity in mind.
Instead of teaching it how to hack, Tencent focused on "defensive" skills, leaving out anything about exploit development or penetration testing.
The goal? Make sure Hy4 helps defend against attacks, not enable them.
Tech industry limits offensive AI features
This move lines up with what other companies like Moonshot AI are doing, limiting their models' offensive abilities as AI-powered cyberattacks become more common.
Both Chinese and US tech giants are rethinking how they train and share these powerful tools, showing a real push toward responsible and safer AI development.