China is proposing a system to label AI-generated content online. The country’s internet watchdog wants to ensure that people know when AI has been involved in creating online media. This proposal covers a wide range of content, from text and images to videos and audio files.
China’s Cyberspace Administration wants to make sure AI content can be easily identified. The draft suggests adding labels to content and its metadata. This is to ensure transparency and raise red flags about unmarked content. If platforms suspect AI-generated content but lack the appropriate metadata, they will be required to label it.
New rules for AI-generated content in audio and video
Under this proposal, AI-generated audio content will need an extra layer of transparency. Not only must the content be labeled at the start and end, but the platforms must also issue voice prompts to alert listeners to AI involvement. If a platform detects AI mid-stream, it may even add another notice at that moment. Similarly, video platforms will need to display notices when AI is detected, with prompts appearing at various points in the video to inform viewers.
Beyond labels, anyone using AI tools on these platforms will be required to identify themselves. The platforms will also have to keep a log of user activity for at least six months. These measures align with China’s ongoing efforts to keep a tight grip on internet activity, ensuring that all online content can be traced back to an individual.
Platforms face pressure under the proposed plan
The proposal includes strict penalties for non-compliance. Service providers that fail to implement these requirements could face important consequences. Beijing wants platforms to be diligent in ensuring that AI-generated content is properly flagged and that anyone using AI to create content is clearly identified. This means that tech giants in China will need to be extra vigilant, as failing to do so could lead to trouble with authorities.
Moreover, platforms are expected to share their methods for detecting AI content, contributing to a broader effort to improve AI detection across the board. This cooperation between platforms is key to Beijing’s vision of tightening control over the internet, ensuring that no content slips through unnoticed.
China usually doesn’t change major regulations once they’re proposed. Since the country is focused on controlling user-generated content, it’s likely that these rules will be implemented with few adjustments. This is another step in China’s ongoing effort to maintain sovereignty over its digital space.
Featured image credit: Furkan Demirkaya / Dall-E
