China's AI human-like interaction rules take effect, banning minors' virtual partners

China's Interim Measures for the Management of AI Human-Like Interaction Services took effect on July 15, 2026, giving the country its first binding rules for AI companion and emotional-support chatbots.

The measures bar platforms from offering virtual partners or relatives to minors and set new limits on chatbots designed to simulate emotional bonds with users.

What it is

The Measures were published by the Cyberspace Administration of China together with the National Development and Reform Commission and the Ministry of Industry and Information Technology, among other agencies.

They target AI products built to mimic human personality and provide ongoing emotional interaction, while leaving other AI tools out of scope.

What it requires

Providers face an absolute prohibition on offering virtual family members or virtual partners to anyone classified as a minor.

Users under 14 need guardian consent before accessing other human-like interaction services at all.

For users of all ages, chatbots must disclose that they are not human and must send usage-duration reminders every two hours.

Noncompliant companies can face fines of 200,000 yuan (£22,000) or the risk that the government publicly circulates criticism of them.

On the day the rules took effect, ByteDance and Alibaba discontinued their AI companion features on Doubao and Qianwen.

What security leaders should do

Security and product teams operating anthropomorphic chatbots reachable from mainland China should map which features count as ongoing emotional interaction, since that framing, not the underlying model, determines whether the Measures apply. Age-verification, guardian-consent capture, and minor-mode logic need to be built into account creation flows rather than bolted on later, given the flat ban on virtual-partner services for minors. Teams should also treat the two-hour usage reminder and AI-disclosure requirements as baseline UX controls to audit alongside existing content-moderation and data-security obligations under China's generative AI and deep synthesis rules, since the new Measures supplement rather than replace them.