China has released new regulations targeting the fast-growing field of AI-powered humanized or personified interactive services online, with authorities citing mounting risks to national security, public welfare, and user well-being. The interim measures, jointly issued by five government bodies including the Cyberspace Administration of China (CAC), the National Development and Reform Commission, and the Ministry of Public Security, will take effect on July 15.
According to Xinhua, the regulations are formally titled the “Interim Measures for the Administration of Artificial Intelligence Humanized Interactive Services.” A CAC official stated the rules aim to promote the healthy development of such services, safeguard national security and public interests, and protect the legal rights of citizens and organizations.
Officials noted that risks associated with AI humanized interactions have become increasingly apparent, including harm to the physical and mental health of minors, threats to information security, dangers to users’ physical safety, and the reinforcement of ethical biases.
The measures explicitly prohibit AI humanized interactive services from generating content that endangers national security, incites the overthrow of state power or the socialist system, or promotes the splitting of the country. Services are also barred from producing content that attempts to extract state secrets, trade secrets, or personal private information.
Further prohibitions cover content that encourages or glorifies self-harm or suicide, uses verbal violence, or could negatively influence minors by prompting unsafe behavior, extreme emotions, or unhealthy habits. AI services must not manipulate users into making unreasonable decisions that damage their lawful interests.
The regulations also introduce a tiered and categorized oversight framework, outline security management obligations for service providers, and establish protections specifically for minors and the elderly. Requirements for safety assessments, algorithm registration, and the development of AI safety service platforms are also included.
Source: Central News Agency (Taiwan), April 10, 2026
https://www.cna.com.tw/news/acn/202604100306.aspx