Ongoing1 articles
Labeling AI: Korea, China, Japan and Taiwan Against the EU's Dec. 2 Deadline
What has to be labeled when AI makes an image, a video or a text, and how? The EU AI Act's labeling duties have applied since Aug. 2, 2026, and on Dec. 2 the grace period for machine-readable marking on generative AI placed on the EU market before that date ends. With the EU as the reference point, this four-part series compares the rules in Korea, China, Japan and Taiwan.
- ・Part 1 The EU: what ends Dec. 2 and what has already begun
- ・Part 2 What a label looks like: visible and machine-readable marking in China and the EU
- ・Part 3 Who pays when the rules are broken: penalties in Korea and Taiwan
- ・Part 4 Where there is no labeling duty: Japan and a checklist for after Dec. 2
Articles
1
EN
EU Deepfake Labels Already Apply. What Changes Dec. 2
EU AI Act deepfake labeling has applied since Aug. 2. Dec. 2 ends only a narrow marking grace period and starts a conditional ban on non-consensual sexual-deepfake AI.