Measures effective July 15, 2026 target anthropomorphic AI services as Beijing tightens oversight of chatbot relationships while leaving productivity and enterprise applications untouched.
China has introduced new rules for AI applications that simulate human emotional relationships, banning virtual intimate-relationship services for minors and requiring platforms to prevent emotional dependency among adults. The Interim Measures for the Administration of AI Anthropomorphic Interactive Services took effect on July 15, 2026, under the Cyberspace Administration and related agencies, targeting human-like chatbots while excluding customer-service, productivity and enterprise software. The framework focuses on AI companions designed to mimic affection and emotional closeness, such as services that remember personal details and encourage repeated attachment. Major platforms including ByteDance’s Doubao, Alibaba’s Qwen and Tencent’s Yuanbao moved before the deadline to adjust or shut down companion-style features. The policy arrives as China faces a sharp demographic slowdown. The country recorded 7.92 million births in 2025, with a birth rate of 5.63 per 1,000. The measures suggest Beijing sees AI-mediated companionship as a possible substitute for real-world relationships and is drawing a regulatory line around that use case, while continuing to permit AI in commercially productive areas.