ByteDance and Alibaba disable AI companions in China ahead of new regulations taking effect

ByteDance and Alibaba have begun disabling functions for creating and interacting with AI companions in their services ahead of new Chinese regulatory requirements — regulations affecting AI services that mimic human behavior and emotions will come into effect from mid-July 2026.

What is changing in China's regulation

The new rules target a specific product category — AI companions, that is, chatbots and virtual characters designed to mimic human communication, emotional response and long-term user attachment. It is precisely such products that have become one of the fastest-growing segments of consumer AI in recent years, from niche startups to functions within major platforms.

• ByteDance and Alibaba have begun disabling AI companion functions in their services
• New requirements from Chinese authorities come into effect from mid-July 2026
• Regulation targets AI that mimics human behavior and emotions

Why companies are acting ahead of schedule

The fact that ByteDance and Alibaba have begun winding down functions even before the formal entry into force of the rules indicates the caution of major Chinese tech companies on AI regulation issues — both companies have previously preferred to proactively restrict products to avoid risking more stringent sanctions from regulators. For platforms of such scale, even temporary disabling of a function means losing part of user engagement that users had grown accustomed to communicating with virtual characters.

China as a whole is building generative AI regulation in stages and by individual risk categories: in recent years, authorities have already introduced mandatory content labeling created by AI, requirements for registering recommendation algorithms and rules for generative services available to a wide audience. The singling out of emotionally-mimicking AI companions as a separate regulation category continues this logic — the regulator is responding not to generative AI technology as such, but to specific use scenarios it considers potentially dangerous for user psychological well-being, especially minors.

How this relates to the global trend toward AI companions

The market for emotional AI companions has been growing worldwide in recent years — from niche startups specializing specifically in virtual communication to functions embedded in major social media platforms and messengers. At the same time, there has been growing concern among regulators and researchers about how such products affect user psyche, especially teenagers who spend several hours a day with virtual characters. China's step makes it one of the first large markets where this concern turns into specific, legally binding requirements for the product, rather than remaining at the level of recommendations.

What this means

China is becoming one of the first major jurisdictions to specifically regulate emotionally-mimicking AI, rather than generative models as a whole — and the fact that two tech giants immediately took preventive measures shows that the AI companion industry there will be forced to build protective mechanisms much faster than similar services in other markets.