China has proposed new regulations to standardize the labelling of AI-generated synthetic content, aiming to protect the legitimate rights and interests of citizens, legal persons, and other organizations, as well as to safeguard public interests.
On 14 September 2024, the Cyberspace Administration of China (“CAC”) released a draft regulation titled Measures for Labelling Artificial Intelligence Generated Synthetic Content (the “Draft Measures”), aimed at regulating the labelling of AI-generated content. On the same day, CAC also released the draft of a mandatory national standard draft, detailing the labelling methods for AI-generated synthetic content for various media types in production, display, and distribution.
Key Points of the Draft Measures include:
1. Scope of Application
The Draft Measures apply to network information service providers (“service providers”) offering AI-generated content services to the public within China. In particular, service providers cover algorithm recommendation service providers, deep synthesis service providers, and generative artificial intelligence service providers. However, the Draft Measures exempt certain entities if they are developing or using AI content generation technology for internal purposes or research, without offering public-facing services within China.
2. Comprehensive Obligations for Service Providers and Platforms
The Draft Measures establish comprehensive obligations for service providers and content distribution platforms. For all service providers, they are required to clearly specify in the user service agreement the method, style and other specifications for labelling of generated synthetic content, and remind the user to carefully read and understand the relevant labelling management requirements. If the user requests the service provider to provide generated synthetic content without explicit labels, such content can be provided after the user agreement specifies the user’s labelling obligations and liabilities for usage, and the relevant logs shall be retained for no less than six months. Specifically, for deep synthesis service providers, they must add explicit labels to AI-generated content across various media types and include implicit labels in file data, ensuring these labels persist in downloaded or exported content. Content distribution platforms are required to verify implicit labels, add visible indicators for AI-generated content, and provide tools for users to declare such content.
3. Compliance Requirements and Enforcement
To ensure compliance, all service providers must adhere to national labelling standards, and shall also submit AI-generated content labelling materials when filing for algorithms and conducting security assessments. The Draft Measures stipulate that violations resulting in severe consequences may incur penalties under applicable laws, underlining the importance of adherence to these new regulations.
4. Implementation and Significance
Based on the text of the current Draft Measures, they are set to take effect in 2024, with the exact date to be determined. This regulation, along with the accompanying national standard, represents a significant step in China’s efforts to manage AI-generated content and its potential impacts on society.
With the growing use of AI to generate content, the associated risks of such content such as misinformation, deepfakes, and IP issues, have the potential to infringe individual’s rights and also undermine social stability. The Draft Measures aim to address concerns about misinformation, security risks, and public trust in content created using artificial intelligence.
The Draft Measures are open for public feedback until 15 October 2024, while the mandatory national standard draft is open for public feedback until 13 November 2024.
The full text of the Draft Measures can be found here, and the mandatory national standard draft here. (Chinese only).
For more information on the investment policies in health care sectors in China, contact your CMS client partner or these CMS experts.
Social Media cookies collect information about you sharing information from our website via social media tools, or analytics to understand your browsing between social media tools or our Social Media campaigns and our own websites. We do this to optimise the mix of channels to provide you with our content. Details concerning the tools in use are in our Privacy Notice.