China has introduced new regulations to control "anthropomorphic AI interactive services," targeting AI that provides continuous emotional interaction by simulating human personality traits, thought patterns, and communication styles. The rules, issued by the Cyberspace Administration and other government agencies, came into effect on July 15.
PolicyCyberspace Administration of China
China Implements Strict Regulations on Anthropomorphic AI to Prevent Emotional Dependence
The regulations require AI providers to ensure their products do not foster emotional dependence, encourage harmful behavior, excessively cater to users, or displace human-to-human relationships. Specific safeguards include mandatory reminders every two hours for adults to remind them that the AI is not a person, and a total ban on "virtual intimate relationships"—such as AI boyfriends or girlfriends—for users under 18.
In response to the looming rules, major providers including Alibaba, Bytedance, and Tencent have cut off users' ability to tailor chatbots to act as companions. Meanwhile, other companies continuing to offer companion AI have implemented age-verification checks and additional guardrails.
The regulatory approach in China is noted for being particularly comprehensive compared to other regions. While the EU focuses on preventing deception or manipulation, Chinese regulations also explicitly forbid AI from generating content that "spreads rumors," incites "subversion of state power or the overthrow of the socialist system," or endangers national honor. The rules do not apply to customer service bots, work assistants, or educational AIs, which the state assumes do not create ongoing emotional connections.
Sources
- China's Regulators Take Aim at "AI Boyfriends" （Hacker News Frontpage, 2026-09-14）