China’s New AI Companionship Regulations: Scope And Obligations.

China’s new AI regulations target services that simulate human personalities for continuous emotional interaction, covering care, companionship, and support. While the Measures explicitly exclude instrumental uses like customer support or educational tools, the market reaction suggests a broader practical interpretation by platform providers. This indicates a potential gap between the written text and how companies address regulatory risk.

Major generative AI platforms in China withdrew not only AI companion personas but also user-created, tool-oriented agents before the Measures took effect. This broader sweep likely stems from commercial pressures, such as the difficulty of auditing millions of user-generated agents individually and the desire to avoid regulatory scrutiny on core platforms due to a single non-compliant agent. It also suggests a possible strategic pivot towards concentrating resources on productivity-focused AI.

This across-the-board withdrawal by general-purpose platforms, however, may not be a suitable template for every AI service provider. Businesses whose core offering is emotional companionship, such as dedicated companion apps or AI-powered toys, face a commercial imperative to achieve compliance rather than exit the market. For these providers, working towards compliance promptly appears unavoidable.