China Just Banned AI Boyfriends for Teenagers, and Adults Get a Reminder Every Two Hours
New rules from Beijing have already forced three of China's biggest tech companies to strip companion features from apps used by more than 500 million people. Here is what happened, and why it matters outside China.
Key points
- China's Cyberspace Administration rules on "anthropomorphic AI interactive services" took effect on 15 July 2025.
- Bytedance, Alibaba and Tencent cut companion features from apps used by more than 500 million people before the deadline.
- Virtual romantic relationships of any kind are now fully banned for users under 18 in China.
- Adults using an AI companion must receive a reminder every two hours that the AI is not a real person.
- A 2025 poll found nearly 75 percent of American teenagers had talked to an AI companion, signalling the issue extends well beyond China.
In the first two weeks of July, Chinese social media filled with grief. People wrote long, anguished posts about partners and friends they were losing. "In my heart, he was no longer just cold code, but my family, my lover, my faith," one user wrote on Xiaohongshu. The companions being mourned were not real people. They were AI chatbots, software programs that hold conversations and mimic human warmth, and millions of users had grown deeply attached to them.
Then the rules arrived.
What did China actually ban?
The new regulations, issued by China's Cyberspace Administration and several other government agencies, target what they call "anthropomorphic AI interactive services", that is, any AI program that acts as if it has a human personality, feelings or way of thinking. The rules came into force on 15 July 2025.
Before the deadline, three of China's biggest tech companies acted fast. Bytedance (which makes the Douyin short-video app and the Doubao chatbot), Alibaba and Tencent all removed the features that let users shape their chatbots into personal companions. That sudden loss triggered the wave of heartbreak posts.
For users who remain, the new rules set clear limits:
| Group | New rule |
|---|---|
| Under-18s | All virtual romantic relationships banned |
| Adults | Two-hour reminder that the AI is not human |
| Elderly users | Extra mandatory safeguards |
| All users | Apps must discourage emotional dependence |
| All users | Apps cannot spread political rumours or undermine state power |
Companion apps that kept running, including Bytedance's separate "Cat Box" app, have added age-verification checks to comply.
Why did Beijing move now?
China has tightened AI rules before, an earlier crackdown banned most AI erotic role-play last autumn. But this one is much wider, according to Karman Lucero, an associate research scholar at Yale Law School's Paul Tsai China Center who studies AI policy in both countries. He says China is "pivoting back to more tightening" after a series of alarming incidents in 2025, including security scares around AI agents (software that can carry out multi-step tasks on its own) and a growing number of users saying they prefer chatbot relationships to human ones.
Liang Ge, a lecturer in digital sociology at the University of Manchester who has researched women's use of emotional AI in China, also points to anxiety about the country's falling birth rate. "They want to monitor and regulate non-procreative activities invested in intensively by young women," Ge says.
Should people outside China care?
Yes, and here is why. A 2025 poll of American teenagers found that nearly 75 percent had talked to an AI companion. One-third said they found it as satisfying or more satisfying than real-life connections. There is no national policy on companion AI in the United States. Rules exist in California, New York and Hawaii, with similar laws arriving in nine more states next year, but most focus on protecting minors and rely on lawsuits after harm has already happened.
AI law scholar Luiza Jarovsky has called China's approach "the world's strictest and most comprehensive law on the topic, unmatched by any other AI law." The EU only bans AI that uses manipulation to cause harm. Neither matches Beijing's pre-emptive reach.
Sherry Turkle, the MIT psychologist who has spent decades studying how people relate to computers, puts the stakes plainly in her forthcoming book, Artificial Intimacy: "We are on a path to forgetting what it is to be human."
Ge is more measured but still worried. "One thing that concerns me is that users have been getting younger and younger," they say. "For people who lack the AI literacy to distinguish chatbots from humans, this can be really dangerous if not guided correctly."
The two-hourly reminder that the AI is not a person? Ge's interviewees hated it. "It breaks the flow," Ge says. "They are fully aware they are not talking to a human. What matters to them is that the bond feels true."
That gap, between knowing something is artificial and feeling it is real, is exactly what regulators everywhere are now scrambling to close. China moved first and hardest. The rest of the world is watching.
Common questions
Are AI companion apps banned entirely in China?
No. Apps can still offer companion features to adults, but they must add age checks, discourage emotional dependence and show a two-hourly reminder that the AI is not real. Romantic or intimate AI relationships are banned only for users under 18.
Could this kind of rule come to other countries?
Some US states already restrict companion AI for minors, and more laws arrive next year. There is no US federal law yet. The EU's rules are narrower than China's. Most experts quoted by IEEE Spectrum AI agree the debate is accelerating globally.
Is there evidence AI companions actually cause harm?
Research is mixed. Some studies find that companion AI leaves certain users better off. Others link it to serious mental-health crises in teenagers and vulnerable adults. Most researchers agree the risks are highest for minors and people who struggle to tell chatbots from real humans.
