CHINA BANS AI ROMANCE CHATBOTS FOR MINORS
China’s new rules on AI romance chatbots took effect on July 15, 2026, banning companies from offering virtual partners or virtual family members to minors and ordering them to stop their bots from fostering emotional dependence. The Interim Measures for the Management of AI Human-Like Interaction Services were issued by five government bodies, led by China’s cyberspace authority.
The Cyberspace Administration approved the measures on February 2, 2026, and published them on April 10 with the National Development and Reform Commission, the Ministry of Industry and Information Technology, the Ministry of Public Security and the State Administration for Market Regulation. That gave the industry three months to get its digital sweethearts in order.
The rules cover services that use AI to simulate human personality traits, thought patterns and communication styles for sustained emotional interaction, Latham & Watkins said. That includes bots built as friends or romantic partners. Customer-service, education, work-assistant and research bots are exempt as long as they do not offer sustained emotional interaction.
The Wall Street Journal summed up the bigger picture: “Beyond the bans on emotional reliance and minors, the new rules require companion chatbots to undergo regulatory evaluation before being offered to the public, and give the government a freer hand to shut down any deemed unsafe.”
The rules for kids, and for everyone else
Children under 14 need a parent’s or guardian’s consent to use any other covered service. Providers must build a minor mode with periodic reminders and usage-time limits, and give parents tools to get safety alerts, see usage summaries, block specific characters and cap spending.
Companies also have to identify minors and move them into that mode, with an appeal channel for anyone flagged by mistake.
For adult users, the measures bar bots from excessively flattering people, inducing addiction or undermining real-world relationships. Chatbots must make clear they are not human, and the BBC reported that they must remind users about time spent online every two hours of continuous use. If a user says they intend self-harm or face a major financial loss, providers must step in and contact a guardian or emergency contact.
Quitting is covered too. When a user asks to stop by interface, voice or keyword, the bot has to comply and cannot keep chatting to keep them from leaving.
Services that launch, add companion features or reach 1 million registered users or 100,000 monthly active users must file safety assessments with their provincial cyberspace authority. Violations can bring warnings, public criticism and suspension of service. Fines run from 10,000 to 100,000 yuan, rising to 100,000 to 200,000 yuan where life or health is endangered and harm results.
Tech giants pull the plug
The industry moved before the deadline. Caixin reported on July 6 that Alibaba’s Qianwen and ByteDance’s Doubao planned to drop their features for creating and chatting with customizable AI characters on July 15. The two apps had monthly user counts of 166 million and 345 million, respectively, and task-oriented agents were set to stay. A later Mint Research newsletter said Tencent had withdrawn comparable features too.
The Wall Street Journal called Beijing’s approach more heavy-handed than U.S. lawmakers’ approach to AI mental-health safeguards.
Regulators were already out checking. Shanghai cyberspace authorities said they ran 16 AI safety-compliance training sessions in local communities and directed nearly 100 key platforms to improve compliance, according to a regulator update. They said they had inspected for sham minor modes, ineffective identity checks and harmful generated content, and had taken enforcement action against some apps.
The rules arrived amid falling marriage and birth rates, Bloomberg reported. Yahoo News / Digital Trends reported that China’s population shrank for the fourth consecutive year in 2025 and that births hit a record low.
That report pointed to AI characters portrayed as having children with human companions, and to AI recreations of deceased family members. It also said concerns have been raised in China that unchecked AI could affect employment, education and fertility.
Ma, a fellow with ERA, a British nonprofit talent and research program, said the rules struck a middle path. “This regulation tried to strike a balance between the AI service providers, the users, as well as guardians of minors using the services,” Ma said, the Economic Times reported.
The Economic Times reported on October 15, 2026 that the rules also require chatbot platforms to warn users against excessive reliance and to intervene when users appear to be in extreme distress.
