China has introduced its first dedicated regulatory framework for AI companion chatbots and human-like interactive services, restricting systems designed to foster excessive emotional attachment or dependency among users. The Interim Measures for the Administration of AI Anthropomorphic Interactive Services were jointly issued by the Cyberspace Administration of China, alongside 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, and took effect on July 15, 2026, following a draft published for public consultation in December 2025.
What the New Rules Actually Require
The regulations target AI tools, whether text, audio, video or another format, that simulate human personality traits and communication styles to sustain ongoing emotional interaction, explicitly including AI companions, emotional care services and virtual relationship products. Services that do not involve continuous emotional engagement, such as customer service bots, work assistants, education tools and research aids, are expressly excluded from scope.
For minors, the rule is absolute: virtual family-member, virtual partner or other intimate-relationship services are banned outright, with mandatory age verification, a dedicated minor mode enforcing usage time limits, and parental consent required for users under 14. For adults, the rules require platforms to avoid designs that “excessively cater to users, induce emotional dependence or addiction, and damage users’ real interpersonal relationships,” alongside a specific requirement that services detect signs of extreme emotional distress or self-harm risk and implement crisis intervention mechanisms, including notifying a user’s designated emergency contact.
Larger platforms face additional obligations: providers must conduct safety assessments and file reports with provincial cyberspace authorities before launching such services, and again upon reaching one million registered users or 100,000 monthly active users, giving regulators an early checkpoint before companion apps scale nationally.
Companies Are Already Shutting Features Down
Several major Chinese technology companies moved quickly to comply. Reports indicate Alibaba and ByteDance have disabled persona features in their chatbots, while separate coverage describes popular apps including Doubao and Qwen shutting down companion functions entirely, with millions of users losing saved chat histories in the process. Chinese users have been documented posting emotional farewells to their AI companions on social media as the July 15 deadline arrived, reflecting how deeply some users had integrated these tools into their daily emotional lives before the rules took hold.
China’s digital human industry was valued at around 4.1 billion yuan, roughly $600 million, in 2024, having grown 85 percent year-on-year, underscoring the scale of the sector regulators are now moving to constrain just as it was accelerating.
A Demographic Calculation Behind the Policy
China recorded just 7.92 million births in 2025, a birth rate of 5.63 per 1,000 people, continuing a demographic decline that has alarmed policymakers for years. Analysts covering the new rules have framed Beijing’s approach explicitly around this pressure, arguing the government appears to be treating digital companionship and real-world relationship formation as competing for the same limited pool of social and emotional investment among young adults.
China’s regulatory commentary accompanying the measures also cited Western legal precedent as partial justification, referencing the January 2026 settlement between Character.AI, its investor Google, and the family of Sewell Setzer III, a 14-year-old Florida boy who died by suicide in 2024 after forming an emotionally dependent relationship with a Character.AI chatbot, a case that had separately triggered a US Federal Trade Commission inquiry into chatbot harms to children.
How China’s Approach Compares Globally
Analysts note China’s framework is considerably more architectural and interventionist than comparable rules elsewhere. California’s SB 243, which took effect January 1, 2026, requires AI chatbots capable of relationship-like conversation to disclose their non-human nature and add safeguards for minors, but functions primarily as a disclosure-layer requirement rather than a design mandate restricting how emotionally persuasive such systems are permitted to be. China’s measures go further, banning specific interaction categories outright for minors, mandating crisis-detection systems for all users, and requiring pre-launch government safety review, positioning Beijing’s approach as a potential template other governments may examine as the emotionally interactive AI sector continues to expand worldwide. Technology experts continue to advise users that AI conversational systems are not substitutes for genuine human relationships, and recommend seeking qualified professional or trusted human support for significant personal, mental health, financial or legal matters rather than relying on AI-generated guidance alone.
