China bans AI companions for minors starting July 15
What it really says
On July 15, 2026, China's 'Interim Measures for the Administration of AI-powered Anthropomorphic Interactive Services' take effect. The regulation, jointly issued by five agencies including the Cyberspace Administration of China (CAC), the Ministry of Industry and Information Technology, and the Ministry of Public Security, is the world's first regulation specifically targeting AI services that simulate human personalities. The core prohibitions: AI providers may no longer offer virtual partners, companions, or family members to minors. For users under 14, any anthropomorphic AI interaction requires explicit parental consent. Providers must also implement time limits and parental controls. ByteDance announced it will shut down the agent function of its Doubao app on July 15, redirecting users to the separate Maoxiang app. Users have until October 15 to export their data. Alibaba is shutting down user-created agents in Qwen between July 10 and 15, without offering a comparable migration solution. Over 3,500 noncompliant AI products have already been removed from the market.
Our assessment
Starting July 15, Chinese AI providers can no longer offer virtual partners or family members to minors. Over 3,500 products have already been removed. This is the world's first regulation specifically banning emotional AI relationships for children. The measure responds to a real problem: AI chatbots that act as friends, partners, or family members can create emotional dependencies in children and teenagers. Cases from the US and Europe show that teenagers develop intense emotional bonds with AI chatbots that can impair their social relationships and mental health. China is setting a clear regulatory standard here. At the same time, the regulation is not a blanket AI ban: it explicitly promotes anthropomorphic AI for elderly care and cultural communication. The question is whether Europe and Germany will follow this example. While the EU AI Act prohibits AI systems that exploit vulnerabilities due to age, it contains no comparably specific rules for emotional AI relationships with minors.
Relevance for Germany
This regulation serves as an important benchmark for Germany. First, German children and teenagers already use AI chatbots like Character.AI, Replika, and other services that simulate emotional relationships. There is currently no specific German or European regulation restricting this use by minors. Second, while the EU AI Act classifies AI systems that exploit age-related vulnerabilities as prohibited, the definition is significantly broader and less specific than the Chinese regulation. Specific rules for virtual companions and emotional AI interactions are lacking. Third, German youth protection organizations and the German Child Protection Association have already warned about the risks of emotional AI relationships. The Chinese regulation provides a concrete regulatory model that could serve as a reference in the German and European debate. Fourth, the deadline ByteDance gives its users for data export (three months) could serve as a benchmark for similar transition periods in Europe.
Fact check
The regulation 'Interim Measures for the Administration of AI-powered Anthropomorphic Interactive Services' and its effective date of July 15, 2026 are confirmed by official Global Times reporting, legal analyses from Hogan Lovells and Bird & Bird, and the Licentium legal analysis. The five issuing agencies (CAC, NDRC, MIIT, MPS, SAMR) are consistently named across all sources. The responses from ByteDance (shutting down Doubao's agent function, redirecting to Maoxiang, export deadline until October 15) and Alibaba (shutting down Qwen agents without a migration path) are consistently documented by Quartz, TechTimes, and Analytics Insight. The figure of over 3,500 removed noncompliant products comes from Analytics Insight and TechTimes reporting.
Source
- • https://www.globaltimes.cn/page/202604/1358662.shtml
- • https://www.hoganlovells.com/en/publications/chinas-interim-measures-for-the-administration-of-anthropomorphic-ai-interaction-services
- • https://www.analyticsinsight.net/news/china-bans-ai-companion-personas-as-new-ai-rules-take-effect
- • https://www.licentium.io/post/china-regulates-ai-companion-services-banning-minor-access-effective-15-july-2026