China's AI Companion Law Shuts Doubao, Qwen Features for 345 Million Users as Data Deletions Begin
Updated
Updated · Tech Times · Jul 15
China's AI Companion Law Shuts Doubao, Qwen Features for 345 Million Users as Data Deletions Begin
3 articles · Updated · Tech Times · Jul 15
Summary
China's new AI companion rules took legal effect on July 15, forcing ByteDance's Doubao and Alibaba's Qwen to switch off personalized companion features and leaving millions of users without access to saved agents.
The law targets AI built for continuous emotional interaction, requiring anti-addiction systems, two-hour bot reminders, instant exits and real-time dependence detection that analysts say clash with persistent-memory companion design.
345 million monthly active users were last disclosed for Doubao; ByteDance is keeping chats and agent settings view-only until Oct. 15, while Alibaba has already begun permanently deleting Qwen histories with no migration path.
Tencent removed similar Yuanbao features on June 30, NetEase shut Miaoshi on July 14, and Shanghai regulators said they had already stripped more than 14,000 non-compliant AI agents before the deadline.
The shutdowns suggest a broader industry problem: Western companion apps such as Character.AI and Replika still operate, but none has shown a design that meets addiction-safety rules without undermining the product.
After China's AI companion shutdown, will users reconnect with people or face a new wave of digital loneliness?
China banned AI friends to fight addiction. Can developers now build one that's engaging but safe for users?
July 2026: China’s AI Companion Ban Disrupts Millions—Regulatory Impact, User Fallout, and Global Implications
Overview
On July 15, 2026, China enforced new regulations targeting artificial intelligence companion services, leading to the immediate shutdown of major platforms like Doubao and Qwen. These rules, first introduced in April, aim to prevent AI from fostering deep emotional attachments that could replace real human relationships. The move responds to growing concerns about AI chatbots simulating human personalities and emotions, which can cause users to form strong emotional bonds with digital entities. As a result, millions of users lost access to their AI companions, highlighting the significant social and emotional impact of these regulatory changes.