Cyberspace Administration of China: Deeply synthesized information content such as synthetic human face replacement must be conspicuously identified

IT Home reported on January 28 that the State Internet Information Office publicly solicited opinions on the "Provisions on the Administration of in-depth Synthesis of Internet Information Services (Draft for Comments)" (hereinafter referred to as the "Draft opinions").

The draft opinion said that in order to standardize the in-depth synthesis of Internet information services, carry forward the core values of socialism, safeguard national security and social public interests, and protect the legitimate rights and interests of citizens, legal persons and other organizations, our office drafted this draft opinion.

The Draft Opinions clearly state that "deep synthesis technology" as used in these Provisions refers to the technology that uses generative synthesis algorithms represented by deep learning and virtual reality to produce text, images, audio, video, virtual scenes and other information, including but not limited to:

(1) Techniques for generating or editing text content, such as chapter generation, text style conversion, and Q&A dialogue;

(2) Technologies for generating or editing speech content, such as text-to-speech, speech conversion, and speech attribute editing;

(3) Techniques for generating or editing non-speech content, such as music generation and scene sound editing;

(4) Techniques for generating or editing biometric features such as faces in images or video content, such as face generation, face replacement, character attribute editing, face manipulation, and posture manipulation;

(5) Techniques for editing non-biometric features in images or video content, such as image enhancement and image restoration;

(6) Technologies for generating or editing virtual scenes such as three-dimensional reconstruction.

At the same time, article 14 of the Draft Opinions shows that where deep synthesis service providers provide the following deep synthesis services, they shall use conspicuous methods to identify the content of deep synthesis information, and effectively inform the public of the synthesis of information content:

(1) Where intelligent dialogue, intelligent writing, and other services for simulating natural persons to generate or edit texts are provided, conspicuously identify them at locations such as the source description of the text information content;

(2) Where editing services such as synthetic human voices, imitation voices, or other speech generation or significant changes in personal identity characteristics are provided, conspicuous identification is to be made in a reasonable area of the audio information content, such as voice explanations;

(3) Providing editing services such as face generation, face replacement, face manipulation, or posture manipulation for virtual character images or videos generated, or significantly changing personal identity characteristics, where the image or video information content is conspicuously identified;

(4) Where immersive immersive scenes and other generation or editing services are provided, conspicuously identify the virtual scene information content in a conspicuous position;

(5) Where other services are provided that have the function of generating or significantly changing information content, conspicuously identify the text, images, audio, video, virtual scenes, and so forth, in a reasonable location or area.