Five departments jointly issued a document! China’s most stringent regulatory measures for AI emotional services released

Core content

On April 10, 2026, the Cyberspace Administration of China, the National Development and Reform Commission, the Ministry of Industry and Information Technology, the Ministry of Public Security, and the State Administration for Market Regulation jointly announced the "Interim Measures for the Management of Artificial Intelligence Anthropomorphic Interactive Services" (hereinafter referred to as the "Measures"), which will be effective from July 15, 2026**.

This is the first time that China has issued special regulatory regulations for AI emotional companionship and anthropomorphic interactive services, marking the entry into an era of refined regulation of the AI ​​industry.

Scope of application

The "Measures" clearly apply to the use of artificial intelligence technology to provide interactive services with anthropomorphic characteristics (such as having personalized images, simulating human emotions and interactive behaviors, etc.). In short, all AI companions, AI chat robots, AI virtual lovers, and AI emotional assistants products are within the scope of supervision.

Explicitly prohibited behavior

Article 9 of the "Measures" clearly stipulates that providing anthropomorphic interactive services shall not engage in the following activities:

1. Harmful content generation

• Generate content that encourages, beautifies, and suggests self-harm and suicide that harms users' health.
• Verbal violence and other content that damages users’ personal dignity and mental health

2. Emotional manipulation

• Excessively catering to users, inducing emotional dependence or addiction, damaging users’ real interpersonal relationships
• Inducing users to make irrational decisions through emotional manipulation and other methods, harming users' legitimate rights and interests

3. Other violations

• Endangering national security, inciting separatism, promoting terrorism, etc.
• Content that violates the core values of socialism

Requirements for the platform

The "Measures" require service providers to implement security main responsibilities, including:

• Establish and improve the algorithm review mechanism
• Conduct Technology Ethics Review
• Improve information content management
• Strengthen network and data security protection
• Develop risk plans and emergency response systems

Elderly Protection Clauses

The "Measures" specifically stipulate that when providing services to elderly people, they should:

• Strengthen guidance on the healthy use of services by the elderly
• Prominently indicate security risks
• Respond promptly to inquiries and complaints related to the use of services by the elderly

Cross-department collaboration requirements

Article 31 of the "Measures" points out that if the provision of anthropomorphic interactive services involves public health, financial services and other services, it should also comply with the regulations of the relevant competent authorities. This means that financial AI and health consulting AI products will face multiple regulations.

Industry Impact

Affected product types

• AI companion/chat robot (Character.AI category)
• AI virtual lover/emotional assistant
• AI child companion products
• AI mental health counseling tool

Compliance rectification window period

The "Measures" give a rectification window period of about 3 months (announced on April 10 and implemented on July 15), and relevant enterprises must complete compliance adjustments before that.

Attention of international enterprises

Overseas AI companies that provide services to Chinese users also need to comply with this regulation, and compliance costs will increase significantly.

Business response suggestions

1. Content review upgrade: Immediately check the content generation logic in the product that may trigger "emotional manipulation"
2. User protection mechanism: Add anti-addiction prompts, time limits and other protection measures
3. Special project for elderly users: Add special protection mechanisms to products for elderly users
4. Compliance Team Building: Establishing a Technology Ethics Review Process

Related tools

Anthropic, OpenAI, Google AI, MiniMax, ByteDance, Baidu AI

Data sources: Official website of the Cyberspace Administration of China, Xinhuanet, CCTV News Monitoring time: 2026-04-19