Particle.news
Download on the App Store
6 ARTICLES
·
2w ago

China Enforces AI Labels as Major Platforms Comply and DeepSeek Mandates Two‑Layer Marking

Officials describe the mandate as a curb on deepfakes, online fraud, copyright abuse, with a national‑security focus.

Image
Chinese AI social media laws
Image
Deepfake technology – which uses AI to manipulate images, audio and video – threatens both individual and national security, according to Chinese regulators. Photo: Shutterstock

Overview

  • The law that took effect this week requires visible labels and embedded identifiers for AI-generated text, images, audio and video across Chinese platforms.
  • WeChat instructs users to declare AI content, forbids removing or hiding system-applied tags, and warns against using AI for false or illegal material.
  • Douyin urges labels on posts containing AI material and says it can detect origins via metadata, while Weibo added a tool to report unlabelled AI content.
  • AI firm DeepSeek will label all outputs on its platform in China with clear on-screen markers plus hidden metadata and prohibits altering those labels, noting possible legal consequences.
  • The rule was drafted by the CAC with MIIT, MPS and NRTA and follows a regulatory campaign targeting AI services, with broader international uptake and technical limits of watermarking still uncertain.