China Bans AI Companions for Minors, Mandates Usage Limits

China implements world's strictest AI companion regulations

China's Cyberspace Administration rolled out comprehensive regulations governing "anthropomorphic AI interactive services" that took effect July 15, 2026. The rules target any AI providing "continuous emotional interaction" by simulating human personality traits, thought patterns, and communication styles.

The crackdown triggered an immediate response from major tech companies. Bytedance, Alibaba, and Tencent—whose general-purpose chatbots serve more than 500 million users—disabled features allowing users to customize bots as companions. The sudden changes prompted waves of distressed social media posts from users mourning lost AI relationships. "He has become a bond in my life, rooted deep in my heart, my spiritual pillar," one Bytedance Doubao user wrote, according to the Taipei Times.

Companies offering dedicated AI companion services, including Doubao's separate Cat Box app, installed age-verification systems and other safeguards to comply with the new law.

Why it matters

China's approach represents the most comprehensive government attempt to regulate AI companions before widespread harm occurs, contrasting sharply with reactive frameworks elsewhere. The regulations reveal tensions between promoting AI adoption for economic productivity while constraining technologies perceived to threaten social stability, declining birth rates, and individual mental health. How enforcement unfolds will signal whether governments can effectively balance innovation incentives against emerging risks from emotionally engaging AI.

Core requirements and prohibited uses

The regulations mandate that AI providers prevent their products from creating emotional dependence, encouraging harmful behavior, excessively catering to users, or displacing human relationships. Systems must nudge users away from excessive time spent with AI companions.

For users under 18, all "virtual intimate relationships"—including AI boyfriends, girlfriends, and grandparents—are completely banned. Additional mandatory safeguards apply to elderly users.

Adults interacting with companion AI now receive reminders every two hours that they're conversing with software, not a person. Liang Ge, a lecturer in digital sociology at the University of Manchester who studies emotional AI in China, notes users find the interruptions annoying. "They are fully aware that they are not talking to a human being," Ge explains. "What is important to them is that the bond feels true."

The rules also prohibit content that "spreads rumors," incites "subversion of state power," or endangers national honor—reflecting government concerns about AI's societal and political impacts beyond individual harm.

Global context and enforcement approaches

China's regulations exceed other jurisdictions in scope and strictness, according to AI-law scholar Luiza Jarovsky. The European Union only bans AI using deception or manipulation for harmful purposes. The United States has no federal policy, though California, New York, and Hawaii have enacted rules, with nine more states implementing laws in 2027. Most U.S. regulations emphasize protecting minors through post-harm lawsuits rather than preventive measures.

Karman Lucero, an associate research scholar at Yale Law School's Paul Tsai China Center, notes China's regulations use deliberately vague language, giving authorities discretion to adjust enforcement based on circumstances. "China's approach is to use regulation with a lot of relatively vague provisions, and they figure out what those provisions actually mean in practice after the fact," Lucero says.

The regulations exempt customer service bots, work assistants, and educational AI, assuming these applications don't create ongoing emotional connections.

Underlying social pressures remain unaddressed

Researchers question whether regulations alone can counter the social forces driving AI companion adoption. One Chinese survey found 60 percent of Gen Z respondents had virtual partners. A 2025 U.S. poll showed nearly 75 percent of teenagers had used AI companions, with one-third finding them as satisfying or more satisfying than real relationships.

Ge's interviews with Chinese women revealed reluctance toward marriage and childbearing, with "virtual love forms an alternative path." Some women in their 30s and 40s initially used AI companions to supplement real relationships but eventually felt closer to the AI than to human partners.

"I don't think you're going to solve the problem of the low marriage rate or the low fertility rate by saying, 'you can't have an AI boyfriend,'" Lucero observes.

Despite prohibitions and dangers, Ge believes "human-AI love will evolve and become an important intimate practice in the future."

These details were first reported by IEEE Spectrum.