China’s first regulations on AI companion services have taken effect, setting new limits on chatbots designed to simulate emotional bonds with users and barring platforms from offering virtual partners or relatives to minors.

The rules, jointly issued by the Cyberspace Administration of China and four other government departments in April, came into force July 15. They cover AI products designed to mimic human personalities, emotions, and communication styles.

The measures come after years of growing use of AI companion apps in China, where users have created virtual lovers modeled on TV characters, recreated deceased relatives through customized voices and personalities, and treated chatbots as long-term emotional confidants.

Regulators and state media have warned that such apps carry security and compliance risks, including emotional dependence, sexually suggestive content, and weak protections for minors.

As such, the new regulations prohibit AI apps from generating content for minors that may incite unsafe behavior, trigger extreme emotions, or foster habits that could harm their physical or mental health. They also should not excessively cater to users in ways that induce emotional dependence or addiction and damage their real-life relationships; manipulate users’ emotions to push them toward unreasonable decisions that harm their legitimate rights and interests; and act as virtual relatives or virtual partners to minors.

If a user is believed to be overly reliant on or addicted to a service, the platform must send them a reminder about overuse every two hours.

Zhang Linghan, dean of the Institute of Artificial Intelligence Law at China University of Political Science and Law, told domestic media outlet China News Weekly that the emotional investment offered by anthropomorphic interactive services can easily lead to psychological dependence in users, resulting in a decline in real-world social skills, particularly a weakening of the ability to resolve emotional conflicts and build deep relationships.

The new regulations apply to any services deemed as designed to foster emotional connection and interaction. Other AI chatbot functions, such as intelligent customer service, knowledge-sharing tools, and workplace assistants, are not affected.

The same day the regulations went into effect, tech giants ByteDance and Alibaba discontinued their AI companion features on Doubao and Qianwen, respectively. Qianwen shut its companion service down immediately, while Doubao is offering users a three-month grace period to migrate their data to its sister app, Maoxiang, before it is permanently deleted on Oct. 15.

Tech companies had already begun increasing regulations of their chatbots and companion services, primarily due to findings of suggestive content in their chats. On June 30, Tencent removed its AI companion service from its chatbot Yuanbao.

In April 2023, leading Chinese AI model developer MiniMax removed its AI companion product, Glow, from multiple app stores due to pornographic content. Its product manager told domestic media that “80% of users would engage in suggestive or sexually explicit conversations with the AI companions they created within Glow.”

In June 2025, AI chat app Zhumengdao was summoned to appear before the Shanghai Municipal Cyberspace Administration for generating vulgar and suggestive content that harmed minors.

After the new regulations took effect, several other AI chatbots reported an influx of users migrating from Doubao and Qianwen, and began requiring ID verification before use.

Editor: Marianne Gunnarsson.

(Header image: Malte Mueller/Getty Creative/VCG)