Notice on the Public Solicitation of Opinions on the "Method for Identifying Artificial Intelligence Generated Synthetic Content (Draft for Soliciting Opinions)" by the Cyberspace Administration of China

In order to standardize the labeling of artificially intelligent generated synthetic content, maintain national security and public interests, protect the legitimate rights and interests of citizens, legal persons, and other organizations, and in accordance with the "People's Republic of China Network Security Law", "Regulations on the Management of Algorithm Recommendations for Internet Information Services", "Regulations on the Management of Deep Synthetic Internet Information Services", and "Interim Measures for the Management of Generative Artificial Intelligence Services", the State Internet Information Office has drafted the "Artificial Intelligence Generated Synthetic Content Labeling Measures (Draft for Comments)" and is now soliciting public feedback. The public can provide feedback through the following methods: 1. Send feedback via email to: biaoshi@cac.gov.cn. 2. Send feedback via mail to: Network Management Technical Bureau of the State Internet Information Office, No. 11 Chegongzhuang Street, Xicheng District, Beijing 100044, and indicate "Draft for Comments on the Labeling of Artificial Intelligence Generated Synthetic Content" on the envelope. The deadline for feedback is October 14, 2024. State Internet Information Office September 14, 2024 Artificial Intelligence Generated Synthetic Content Labeling Measures (Draft for Comments) Article 1 In order to promote the healthy development of artificial intelligence, standardize the labeling of artificially intelligent generated synthetic content, protect the legitimate rights and interests of citizens, legal persons, and other organizations, and uphold social public interests, these measures are formulated in accordance with the "People's Republic of China Network Security Law", "Regulations on the Management of Algorithm Recommendations for Internet Information Services", "Regulations on the Management of Deep Synthetic Internet Information Services", "Interim Measures for the Management of Generative Artificial Intelligence Services" and other laws, administrative regulations, and departmental rules. Article 2 These measures apply to network information service providers (hereinafter referred to as "service providers") who engage in the labeling of artificially intelligent generated synthetic content in accordance with the provisions of the "Regulations on the Management of Algorithm Recommendations for Internet Information Services", "Regulations on the Management of Deep Synthetic Internet Information Services", and "Interim Measures for the Management of Generative Artificial Intelligence Services". Industry organizations, enterprises, educational and research institutions, public cultural institutions, and other professional institutions engaged in the research and application of generative artificial intelligence technologies, but not providing services to the domestic public, are not subject to the provisions of these measures. Article 3 Artificially intelligent generated synthetic content refers to text, images, audio, video, and other information created, generated, or synthesized using artificial intelligence technology. The labeling of artificial intelligence-generated synthetic content includes explicit labeling and implicit labeling. Explicit labeling refers to labels added in the generated synthetic content or interactive interface in a way that is clearly perceivable by users, such as through text, sound, graphics, etc. Implicit labeling refers to labels added to the file data of generated synthetic content using technical measures that are not easily perceived by users. Article 4 If the services provided by service providers fall under the circumstances specified in Article 1 of the "Regulations on the Management of Deep Synthetic Internet Information Services", they should add explicit labels to the generated synthetic content in accordance with the following requirements: (1) Add text prompts or universal symbol prompts at appropriate positions at the beginning, end, or middle of the text, or place prominent prompts in the interactive interface or around the text. (2) Add voice prompts or audio rhythm prompts at appropriate positions at the beginning, end, or middle of the audio, or place prominent prompts in the interactive interface. (3) Add prominent prompts at appropriate positions in the image. (4) Add prominent prompts at appropriate positions at the beginning of the video and around the video playback, and add prominent prompts at the end and middle positions of the video when necessary. (5) When presenting virtual scenes, add prominent prompts at appropriate positions at the beginning of the scene, and add prominent prompts at appropriate positions during the ongoing service in the virtual scene. (6) In other scenarios of generative synthetic services, add explicit labels with significant prompt effects according to the characteristics of the application. When providing download, copy, export, and other ways of accessing generated synthetic content, service providers should ensure that the files contain the required explicit labels. Article 5 Service providers should add implicit labels to the metadata of the generated synthetic content in accordance with Article 16 of the "Regulations on the Management of Deep Synthetic Internet Information Services". The implicit labels should include information about the attributes of the generated synthetic content, the name or code of the service provider, content numbers, and other production elements. Service providers are encouraged to add forms of implicit labeling such as digital watermarks to the generated synthetic content. Metadata refers to descriptive information embedded in the header of a file according to a specific encoding format, used to record information about the source, attributes, purpose, copyright, and other content of the file. Article 6 Service providers that provide network information content distribution platform services should take measures to regulate the dissemination of generated synthetic content. (1) They should verify whether there are implicit labels in the file metadata, and if so, they should add prominent prompts around the published content to clearly alert users that the content is generated synthetic content. (2) If no implicit labels are found in the file metadata, but the user declares it as generated synthetic content, appropriate prompts should be added around the published content to remind users that the content may be generated synthetic content. (3) If no implicit labels are found in the file metadata and the user does not declare it as generated synthetic content, but the service provider distributing the network information content detects explicit labels or other traces of generated synthetic content, it can be identified as suspected generated synthetic content, and appropriate prompts should be added around the published content to remind users that the content is suspected to be generated synthetic content. (4) For content that is verified, possibly, or suspected to be generated synthetic content, information such as the attributes of the generated synthetic content, the name or code of the distribution platform, content numbers, etc., should be added to the file metadata. (5) Provide necessary labeling functions and remind users to actively declare whether the published content contains generated synthetic content. Article 7 Internet application distribution platforms should verify whether service providers have provided the necessary labeling functions for generated synthetic content when reviewing applications for release or online distribution. Article 8Service providers should clearly specify in the user service agreement the methods, styles, and other specifications for generating synthetic content identifiers, and remind users to carefully read and understand the relevant requirements for managing identifiers.Article 9 If users need the service provider to provide generated synthetic content without explicit identification, they can provide generated synthetic content without explicit identification and keep related logs for no less than six months after clarifying the user's identification obligations and usage responsibilities through the user agreement. Article 10 When users upload generated synthetic content to the service provider's platform for disseminating network information content services, they should proactively declare and use the identification function provided by the platform for identification. No organization or individual shall maliciously delete, tamper with, forge, or conceal the identification of generated synthetic content stipulated in these measures, provide tools or services for others to carry out the aforementioned malicious acts, or damage the legitimate rights and interests of others through improper identification methods. Article 11 Service providers should identify in accordance with the requirements of relevant mandatory national standards. Article 12 When service providers perform procedures such as algorithm filing and security assessments, they should provide materials related to identifying generated synthetic content according to these measures, strengthen the sharing of identification information, and provide support and assistance to prevent and combat related illegal criminal activities. Article 13 For violations of these measures that result in serious consequences due to the lack of identification of generated synthetic content, the relevant competent authorities such as the Cyberspace Administration of China shall impose penalties in accordance with relevant laws, administrative regulations, and departmental rules. Article 14 These measures shall be implemented from the date of MM/YY. This article is from the WeChat official account "Cyberspace China"