China's Rules For Humanlike AI Chatbots Take Effect
Five agencies, the cyberspace, planning, industry, police and market regulators, signed the measures April 10 to govern AI posing as human.
- Cyberspace, planning, industry, public security and market regulators jointly signed the rules on April 10, 2026.
- Measures ban AI content that endangers national security or incites subversion of the socialist system.
- Providers must register algorithms, complete safety assessments and join a state AI sandbox platform.
Why it matters: The state that must approve a chatbot before it can sound human just claimed the human voice.
Cyberspace Administration of China (CAC) — 人工智能拟人化互动服务管理暂行办法 ↗ · Jul 15, 20267/15/26