China enacts world's strictest rules on AI companions, citing emotional harm

October 1, 2026 at 04:25 AM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

China on July 15 introduced the world's most far-reaching regulations on human-like AI chatbots, restricting any AI capable of sustained emotional interaction with a person. The rules target AI companions and large language models that imitate human interaction, following a public debate sparked by a draft released in late 2025. The move comes as Beijing and Washington agreed to launch a new AI dialogue after a meeting between Xi Jinping and Donald Trump last week.

Key Facts

China's new regulation, effective July 15, places strict limitations on any AI capable of sustained emotional interaction with a human.
Douyin's suicide prevention team reported intervening in 400 attempted suicide cases in the first half of 2025.
Sensor Tower estimates about 128 million monthly active users of AI companion apps outside China, with 29 to 70 million inside China.
The draft regulation released in late 2025 triggered a huge discussion on social media, according to Zilan Qian of the Oxford China Policy Lab.
After the Xi-Trump meeting last week, Beijing said the two sides agreed to launch a new AI dialogue with further talks planned for late November.

Regulatory Scope

The July 15 regulation covers apps specifically designed as AI friends, boyfriends, or girlfriends, as well as any large language model that imitates human interaction. It is the most far-reaching set of rules in the world on human-like AI chatbots, according to the BBC report. The rules were introduced after a draft version was shared by the government at the end of last year. Zilan Qian, an AI researcher at the Oxford China Policy Lab think tank, said the draft triggered the first time he had seen people care so much about regulation.

Surveillance and Intervention

Douyin's suicide prevention team operates 24 hours a day and shares user posts with police stations, as shown in a documentary published on Chinese state media last year. In one case in Huangshi, police geolocated a user's video to a local park and intervened before he could take his own life. The documentary claimed the team successfully intervened in 400 cases of attempted suicide in the first half of 2025. The BBC describes this as a window into the architecture of surveillance China has built to monitor and control technology use.

AI Companion Usage

Sensor Tower, a market intelligence firm, estimates about 128 million monthly active users of AI companion apps outside China. Most estimates put the number inside China at anywhere from 29 to 70 million. The new rules extend government control to the realm of relationships and artificial intelligence. Beijing says it has recognized the power of AI to replace human relationships as an existential risk.