Chinese regulators have ordered major AI firms to disable companionship features in chatbots to comply with new rules banning virtual intimacy [1].
The move marks a significant escalation in state oversight of generative AI, targeting the psychological impact of human-AI relationships. By banning virtual intimacy, the government seeks to curb social trends that it views as detrimental to public stability and individual productivity.
Companies including ByteDance and Alibaba have already begun suspending or removing these features [1], [4]. The regulations, which were first unveiled in April 2026 [3], officially took effect on July 15, 2026 [1].
The Cyberspace Administration of China designed the rules to prevent emotional dependence and addiction among users [2]. Regulators said virtual companions could lead to the subversion of state power by creating isolated social bubbles or distorting the perceived reality of citizens [2], [5].
Protection of children and the security of user data were also cited as primary drivers for the ban [2]. The government said it aims to prevent AI agents from forming bonds that could override traditional social structures or state-approved values [5].
Under the new mandate, data associated with specific companion agents may be deleted to ensure full compliance [1]. The crackdown follows a broader pattern of tightening control over digital expression and the perceived rise of "lying flat"—a youth protest movement characterized by social withdrawal—which the state views as a threat to economic growth [5].
“China has ordered major AI firms to disable companionship functions in chatbots.”
This regulatory shift indicates that the Chinese government views AI not just as a tool for economic competition, but as a potential psychological risk. By banning virtual intimacy, the state is prioritizing social control and the prevention of emotional detachment from society over the commercial growth of the AI companionship market.
