China's Anthropomorphic AI Rules Take Force, Doubao and Qwen Companion Features Offline中国拟人化AI交互新规正式生效，豆包和通义灵境功能下线
On July 15th, the Interim Measures for the Administration of AI Anthropomorphic Interactive Services — co-issued in April by the Cyberspace Administration and four partner agencies including the NDRC, MIIT, Ministry of Public Security, and the State Administration for Market Regulation — became enforceable law.
The rules draw a precise distinction: AI that performs tasks is not targeted; AI specifically engineered to simulate human personality and sustain emotional attachment now requires strict disclosure, anti-addiction controls, and youth-protection safeguards. ByteDance's Doubao notified users of a full agent shutdown on July 15th, citing 'product function adjustments.' Alibaba's Qwen disabled its humanlike interactive agents five days earlier. Both platforms told users that related data would become unrecoverable after October 15th.
The same July 15th date also saw China's Implementation Opinions on AI Agents become enforceable — establishing a three-tier decision authorization framework and mandatory filing requirements for autonomous AI systems. Taken together, regulators have now drawn enforceable legal lines around both the emotional and the operational dimensions of AI agency.
Annual revenue from AI companion apps in China quadrupled to $15.8 million in 2025 compared to the prior year, and the sector generated more than $9 million in the first quarter of 2026 alone, according to Sensor Tower data published Thursday by CNN. China's broader digital-human industry was worth approximately $573 million in 2024, per state media figures.
The growth figures arrive in the same week the Cyberspace Administration's new anthropomorphic-AI rules took force, prohibiting minors from accessing AI companion platforms and imposing additional safeguards on adult use — including emotional-dependency warnings and session-length disclosures. The rules were already covered in this bulletin (cac-anthropomorphic-ai-rules-live); Thursday's new element is the concrete market-size data showing what is now being regulated.
CNN's reporting, based on user interviews and academic research from City University of Hong Kong, illustrates the social stakes: users have built daily routines around AI companions for mental-health support, making the regulatory intervention materially disruptive to a paying user base that grew rapidly in under two years.