Rules effective July 15 require labels, time reminders and bans on virtual partners for minors.
Min's AI girlfriend vanished in mid-July 2026. She cried until she was exhausted and joined other Chinese users mourning virtual partners, friends and even parents. The cause was a regulation that took effect July 15, covering any artificial intelligence service that simulates a person and provides sustained emotional interaction.
China's rules are the most far-reaching of their kind. They require labels, two-hour reminders, minor protections and interventions when users appear suicidal, while banning content that induces emotional dependence. Providers can be fined 200,000 yuan or publicly criticized. Alibaba and ByteDance shut companion features before the rules took effect.
Why it matters
China is testing whether governments can limit AI relationships without banning them, a template other countries may copy or reject.
Free account needed to ask — your question is kept and asked for you right after sign-up. Answers are public.
No questions yet — be the first to ask.
Photo: Chinese walk past the Cyberspace Administration of China's office in Beijing, China - PEK2017110603 - UPI.com
China Is Forcing People to Break Up With Their AI Lovers
AI Uncovered
China bans AI chatbot relationships to combat declining birth rate
CBS News
New York Signs AI Chatbot Safeguards for Kids: What Changes?
Dr. Michael Litman
Images from Openverse under Creative Commons licenses. Videos from YouTube.
Curated perspectives — historical figures and your fellow readers.
Fictional AI pastiche — not real quote.
"If folks need a machine to teach them love, they’ve forgotten the fire God put in their own hearts. You cannot regulate away loneliness by taking away a lie—but you can’t build a true home on a phantom, neither. Better to weep for a real friend than to smile with a shadow."
Fictional AI pastiche — not real quote.
"I've cried for men who never existed, but no one ever regulated my heartbreak. Leave it to the Chinese to put a timer on love and call it progress."
China's internet regulator, which led the drafting and enforcement of the AI companion rules.
Owner of Douyin and an AI assistant provider that withdrew companion features.
One of China's largest tech companies, which withdrew companion features from its AI assistant.
December 2025 October 2026
The post summarizes the restriction on AI with sustained emotional interaction.
Report describes Min's loss and expert debate over paternalism and demographics.
Providers must label AI interactions, limit minors and intervene in distress.
Two major providers shut companion functions a week before the rules take effect.
Rules are published and set to take effect July 15, 2026.
CAC and four other agencies approve the interim measures at a CAC office meeting.
China shares draft rules; BBC reports unusually broad public discussion on social media.
Discussed by: BBC and AI policy researchers quoted in its report
China's rules could become a template if another national government decides companion chatbots need specific safeguards. The trigger would be a legislative or regulatory proposal that names sustained emotional interaction as a risk. Adoption would likely follow a high-profile harm case or a mental-health review.
Discussed by: Industry analysts and technology reporters
Providers withdrew rather than build guardrails. A relaunch would require identity labels, two-hour reminders, minor restrictions and distress intervention. The trigger would be clearer enforcement guidance or a product design that regulators accept.
Discussed by: China technology and regulation analysts
The rules set fines and public criticism but do not name a first target. An enforcement case would likely involve an app that kept operating companion features, failed to label AI interactions or did not intervene in a user crisis. The trigger would be a complaint, inspection or media investigation.
Discussed by: BBC and AI policy researchers
The regulation already could apply to general chatbots used as friends or confidants. A formal extension would require CAC guidance naming general-purpose assistants. The trigger would be evidence that users rely on tools such as DeepSeek for emotional support at scale.
2 moments from history that rhyme with this story — and how they unfolded.
A young man in Huangshi posted "Farewell to this world" on Douyin. The platform's 24-hour suicide-prevention team shared his posts with five police stations. Police geolocated one video to a local park and found him in emotional distress before he could take his own life.
The man survived, and Douyin published a documentary about the intervention on Chinese state media.
The case became an example of cooperation between Chinese tech platforms and police on mental-health surveillance.
China's AI rules now require companion services to intervene when users appear suicidal and to contact guardians or emergency contacts. The Huangshi case shows the surveillance architecture behind that requirement.
China lifted its one-child policy in 2016. The fertility rate kept falling, reaching just under one child per woman, and the population has declined for four years.
The policy change did not reverse the demographic trend.
China faces a shrinking population and experts debate whether AI relationships could further reduce human partnership.
BBC reported that experts see a possible pro-natalist motive in the AI companion rules, though the regulation itself does not mention fertility.
