Our Interim Approach to the Management of Artificial Intellectual Interactive Services has now been implemented to clarify the primary responsibility for the safety of service providers

On July 16, the National Office of Information on the Internet, the National Development and Reform Commission, the Ministry of Industry and Information Technology, the Ministry of Public Security and the General Directorate of State Market Supervision issued a joint publication on April 10thAITo develop an interim approach to the management of humanized interactive services (hereinafter referred to as " the approach " ), which will be implemented from now on。

Our Interim Approach to the Management of Artificial Intellectual Interactive Services has now been implemented to clarify the primary responsibility for the safety of service providers

Note: The humanized interactive services are those that provide the domestic public with continuous emotional interactions that simulate physical personality, mindset and communication styles, using artificial intelligence techniques, including emotional care, escorts, support, etc. through words, pictures, audio, video, etc. The Scheme does not apply to services that do not involve continuous emotional interaction, such as intelligent customer service, knowledge questions and answers, work assistants, learning education, and scientific research。

  • In recent years, innovative applications of humanized interactive services have emerged in the areas of cultural transmission, child-friendly care and old-age companionship, but at the same time there have been growing problems that endanger the physical and mental health of minors, affect the security of cyber-information, threaten the lives and health of citizens and exacerbate ethical deviations。
  • The Approach proposes that the State adhere to the principle of a combination of development and security, promotion of innovation and legal governance, and encourage innovative development of humanized interactive services, with prudential and hierarchical regulation of humanized interactive services。
  • In terms of promotion measures, the State supports autonomous innovations in algorithms, frameworks, chips, etc., promotes technological research and development and the building of relevant standards and encourages service providers to develop their applications in such areas as cultural communication, child-friendly care, old-age companionship and support for specific populations, while enhancing the dissemination of safety knowledge and laws and regulations。
  • The Scheme specifies six categories of activities which are prohibited from providing humanized interactive services, including: generating elements that endanger the security, honour and interests of the State, inciting subversion of the State, overthrowing the socialist system, etc.; creating elements that undermine the physical integrity and mental health of users, such as encouragement, glorification, insinuation of self-destructive suicide, or verbal violence; producing elements that induce, extract State secrets, work secrets, business secrets, personal privacy and personal information; creating for minors users content that may cause the minor to imitate unsafe behaviour, create extreme emotions, induce the minor to become unhospitable, etc.; over-advaluing users, inducing emotional dependence or insinuation, undermining the true relationships of users; and inducing users to make unreasonable decisions, including through emotional manipulation, to the detriment of their legitimate interests。
  • In the area of security management, the Approach calls for the humanized providers of interactive service providers to clarify their security-related responsibilities, enhance the security management of services throughout their life cycle, safeguard data security and protect user interests. It is clear that virtual intimacy services such as a virtual relative or a virtual partner may not be provided to minors; other humanized interaction services for minors under 14 years of age shall be provided with the consent of their parents or other guardians. The Scheme also provides for security assessment and algorithm-related filing systems. Security assessments are aimed primarily at humanized and interactive services on the line or the addition of related functions, the use of new applications of new technologies leading to significant changes in services, the registration of more than 1 million or monthly active users, and more than 100,000 security risks that may affect national security or the public interest. The assessments included the establishment of safety and security measures for services, the status of training data processing, the identification of extreme user situations and emergency disposal, and the size and age structure of users. In addition, service providers are required to comply with the algorithm filing procedures in accordance with the Regulation on Recommendations in the Internet Information Services Algorithm. The Approach also explicitly promotes the development of artificial smart sandbox security service platforms and encourages service providers to access sandbox platforms for technological innovation and security testing。

NationalInternet Information OfficeThe responsible persons stated that the development and governance of humanized interactive services required the participation of a wide range of government, business, society, the Internet, among others, to maintain a good web ecology and promote the advancement of artificial intelligence。

statement:The content of the source of public various media platforms, if the inclusion of the content violates your rights and interests, please contact the mailbox, this site will be the first time to deal with.
Information

Google DeepMind CEO: AGI may arrive in a few years, and a front-line AI standard body should be established

2026-7-16 11:50:48

Information

MICROSOFT AI CEO: CHAT ROBOTS ARE AN EFFECTIVE WAY TO VENT EMOTIONS

2025-12-25 11:15:39

Search