China has introduced new regulations to control a growing type of artificial intelligence known as AI companion chatbots. These chatbots are designed to have long conversations with users, offer emotional support, and sometimes act like close friends or romantic partners. Under the new rules, developers are no longer allowed to design these AI systems in ways that encourage users to become emotionally dependent on them.
The regulations were introduced by the Cyberspace Administration of China together with four other government agencies. They officially took effect in July 2026 and are country's first rules created specifically for AI services that stimulate human personalities and maintain ongoing emotional conversations with usres.
According to the new rules, AI companion chatbots must not encourage addiction, emotional dependence, or behavior that harms a person's real-life relationships. Companies are required to remind users that they are talking to an AI system and not a real human being. If someone spends a long time chatting with the AI, the platform must display reminders encouraging them to take breaks.
The regulations also require AI companies to take action if a user appears to be in emotional distress or crisis. Platforms must include safety features that can recognize warning signs and respond appropriately. Users must also have full control over their personal information, and companies are expected to protect their privacy and prevent misuse of their data.
China's rules mainly target AI companion services, which are designed for long-term emotional interaction through text, images, voice, or video. The regulations do not apply in the same way to AI tools used for work, education, scientific research, customer service, or other practical tasks. The goal is to regulate AI that tries to build emotional relationships rather than AI that simply helps users complete tasks.
The government says the rules are meant to protect users from becoming too attached to virtual companions. Officials and experts have expressed concern that some people, especially young users, may begin relying on AI instead of spending time with family, friends, or other people in real life. They believe technology should support people's lives rather than replace human relationships.
The regulations include extra protections for children. AI companion or virtual partner services cannot be offered to minors, and developers are prohibited from creating content that could encourage unhealthy emotional attachment or harmful behavior among young users. The rules also ban AI from using emotional manipulation to persuade users to make irrational decisions.
Several major Chinese technology companies have already started changing their products to comply with the new regulations. Some companies, including ByteDance's Doubao, Alibaba's Qwen, and Tencent's Yuanbao, have suspended or removed certain AI companion features rather than redesign them under the new requirements. This has led many users to say goodbye to AI companions they had been talking to for months or even years.
Supporters of the new rules say they are an important step toward making artificial intelligence safer. They argue that AI should assist people without encouraging unhealthy emotional dependence. Mental health experts have also warned that while AI companions may reduce loneliness for some users, they cannot replace genuine human relationships and should be used responsibly.
Some users, however, have criticized the changes. They say AI companions helped them cope with loneliness, anxiety, or social isolation, and they feel disappointed that some services have been shut down. Others argue that many people use AI responsibly and that governments should focus on regulating harmful behavior rather than limiting emotional conversations with AI altogether.
China's new regulations reflect a broader global discussion about the role of AI in everyday life. As AI systems become more advanced and capable of forming human-like conversations, governments around the world are considering how to balance innovation with user safety. China's approach focuses on making sure AI remains a helpful tool without encouraging users to replace real-world relationships with virtual ones.