China is ramping up efforts to make artificial intelligence (AI) safer and more reliable by pairing new laws with advanced technical safeguards, according to experts at the just-ended 2026 World Internet Conference (WIC) Wuzhen Summit in east China's Zhejiang Province.
The four-day event wrapped up Thursday and centered on the theme "Open Source and Openness, Joint Contribution and Shared Benefits -- Jointly Building a Community with a Shared Future in Cyberspace."
The event showcased China's long-term strategy for comprehensive AI governance, highlighting a dual approach that pairs rapid tech development with strict risk control. Over the years, China has steadily built up laws, ethical guidelines, and application norms to ensure AI remains safe, reliable, and controllable.
A major foundation of this strategy is legal regulation. China has accelerated cyber lawmaking, introducing over 190 laws and administrative regulations, including landmark legislation such as the Cybersecurity Law, Data Security Law, and Personal Information Protection Law.
Alongside national laws, the operational framework continues to evolve. This past September, China launched its AI Safety Governance Framework 3.0, designed to build industry-wide consensus on managing security risks in real time.
Beyond national laws and regulations, China is also placing heavy emphasis on technical solutions to keep pace with rapidly evolving AI risks. That effort took center stage at the "Light of the Internet" Expo, a flagship exhibition at the 2026 WIC Wuzhen Summit dedicated to cybersecurity and governance.
To tackle real-world threats such as AI-generated deepfakes, Chinese tech platforms are turning large language models into governance tools, essentially using AI to police AI.
Kuashou Technology, a major Chinese short-video platform, is deploying this exact strategy to build a tighter defense network.
"We use an inherent, certain 'white box' to deal with external, uncertain 'black boxes.' We call it a method of using technology to govern technology. We also introduce AI to govern various problems brought by AI. This is our most routine operation. Combined with human judgment and solid input provided by intelligent agents in more professional scenarios, it forms a relatively tight protection network," said Ma Hongbin, senior vice president of Kuashou Technology.
Together, these legal regulations, technical safeguards, and industry practices form China's integrated framework for AI safety.
That collaborative framework is winning support both at home and abroad. According to Xiao Qian, Vice Dean of the Institute for AI International Governance at Tsinghua University, wide consultation has been key to its success.
"Different stakeholders are invited into the consultation mechanism before measures or frameworks are introduced, and everyone gives different opinions, finally forming a consensus. That's why the current AI Safety Governance Framework 3.0 is recognized by so many people, especially internationally, by many peers, because it truly absorbed suggestions from different parties for improvement," said Xiao.
China highlights multi-layered AI governance strategy at World Internet Conference
