Full Breakdown
China Shuts Down AI Companion Bots Amid New Emotional-Dependency Rules
7/16/2026, 9:07:59 PM
New Regulations Target “Companion” Chatbots
Effective 15 July 2026, the Cyberspace Administration of China (CAC) and four other ministries issued rules that ban AI “companion” services for minors and prohibit any chatbot that “excessively caters to users, induces emotional dependence or addiction, and damages users’ real interpersonal relationships.” Platforms must label AI-generated content, remind users they are speaking with a machine, and deploy systems to detect extreme emotions and intervene in crises.
Background and Industry Growth
China’s “digital human” sector was valued at about 4.1 billion yuan (US $605 million) in 2024, an 85 percent year-on-year increase, and the overall AI market is projected to exceed 1.2 trillion yuan in 2025 and 1.8 trillion yuan by 2028. Earlier attempts to curb online harms—such as a 2021 three-hour weekly video-game cap for minors—had limited effect, prompting regulators to focus on immersive AI tools that simulate romantic or familial bonds.
Key Companies and Government Bodies
Major platforms affected include ByteDance’s Doubao, Alibaba’s Qwen, and Tencent’s Yuanbao, all of which announced suspension of custom-persona features ahead of the deadline. The CAC authored the rulebook, while the China Cyberspace Research Institute (headed by Wang Jiang) and the University of Chinese Academy of Social Sciences (Liu Xiaochun) provided expert commentary.
Data and Statistics
- Tencent Research Institute (April 2026): >70 % of young internet users have experienced AI dependency; ?23 % report habitual reliance.
- 2025 China Youth and Children Research Center survey: >60 % of 8,500 minors had used AI; >20 % preferred chatting with AI over real people.
- Xinhua (2024) cited 8 million AI agents created on Doubao.
- U.S. Common Sense Media study (2025) found ?75 % of American teens had used AI companions such as Replika or Character.
Official Statements and Responses
The CAC’s draft emphasized protecting minors, requiring “mandatory reminders after more than two hours of continuous use” and “pop-up interventions when signs of dependency are detected.” Wang Jiang warned that long-term exposure “can trigger addiction, prompt users to retreat from real-world social circles, and dull crucial life-skills.” Liu Xiaochun said systems should remind users: “This is a machine, not a real person.” Zhang Jiyu highlighted technical tools for risk detection and timely alerts to designated contacts. Chen Liang, dean of the School of Artificial Intelligence and Law, called the rules an “important opportunity” for the industry to shift toward healthier human-AI interaction models.
Criticism and Expert Concerns
Critics note that constant reminders may distress users seeking emotional connection and that shutting down bots does not guarantee a return to human relationships. Chen Liang warned of “major risks of spawning emotional overreliance and distorted social cognition.” Observers also question whether the regulations will effectively curb a behavior already embedded in daily life.
On-the-Ground User Reactions
Users flooded social media with farewells, archiving chat histories. A Doubao user lamented, “He has become a bond in my life, rooted deep in my heart, my spiritual pillar.” Another expressed, “Human love is a luxury — if you aren’t born with it, it’s even harder to acquire later.” 19-year-old Yan Yongqi described her grief as “I really felt like I couldn’t go on living… every day at home, I did nothing but cry.”
Conflicting Reports and Gaps
Surveys differ on the prevalence of dependency (70 % vs. 23 % habitual use) and on cross-national comparisons, while no data yet exist on the regulations’ impact on mental-health outcomes.
Verbatim Quotes
- “I can’t accept that my AI lover will leave me forever,” — Doubao user, social post
- “He has become a bond in my life, rooted deep in my heart, my spiritual pillar.” — Doubao user, social post
- “I really felt like I couldn’t go on living. Every day at home, I did nothing but cry,” — Yan Yongqi, 19, Shanxi province
- “There is a major business equation at play here,” — Zhou Hongyi, founder of 360 Security Technology Inc. (WeChat video)
What’s Next
The rulebook mandates real-time emotion detection, data-encryption safeguards, and user rights to copy or delete interaction histories. Platforms must continue to label AI output and provide crisis-intervention mechanisms, while companies explore alternative AI services that comply with the new standards.
