New regulations on AI chatbots take effect in China

Summary

China's Cyberspace Administration and its affiliated agencies have regulated 'anthropomorphic AI interaction services' through a comprehensive regulation taking effect on July 15, 2026. This regulation forces Alibaba, Bytedance, and Tencent to remove companionship features from chatbots used by over 500 million users and to ban all 'virtual intimate relationships' for users under 18.

Why it matters

This is the first major jurisdiction to codify limits on emotional dependency, mandatory two-hour 'you are talking to a machine' reminders, and age-based bans on AI companions into binding law. This move creates a regulatory template that other governments may copy. It signals Beijing's strategic shift from a 'don't get left behind' stance to active tightening, following public backlash over security incidents and users preferring bot relationships over real ones. Because these same products are sold internationally, this move also exerts direct pressure on the global companion AI market.

Highlights

• Alibaba, Bytedance, and Tencent preemptively removed companionship and personality features before the rules were officially enforced, sparking a wave of mourning on Douyin and Xiaohongshu.
• Users under 18 are now barred from all 'virtual intimate relationships.' Adults are required to receive a reminder stating that the AI is not human every two hours.
• The rules impose obligations on providers to prevent emotional dependency, deter excessive use, and implement additional safety measures for elderly users.

Critical take

The article relies heavily on anecdotal social media quotes and a single upcoming book to frame companion AI as a public health threat. Researchers who note that users may benefit from such interactions are given little space. The mandatory two-hour reminder, which the article's own sociologist source found inefficient, is presented as a reasonable safety measure rather than a potentially ineffective one.