GermanyThe article discusses new transparency requirements introduced by the EU, which mandate that AI-generated content be marked with visible and machine-readable watermarks starting from August 2, 2026. These marks aim to make AI content identifiable for both humans and machines. However, current watermarking techniques are not foolproof, as they can be easily removed using available tools. Experts note that while some companies and researchers are developing more robust methods, such as embedding watermarks deeply within image structures, these approaches are still costly and not widely adopted. The article highlights challenges such as data loss during compression and the potential for open-source AI models to bypass these labels.
Bias read (Center): The article presents a balanced overview of the EU’s new regulations on AI content labeling, discussing both the intent behind the policy and the technical limitations of current watermarking methods. It cites expert opinions from both industry representatives and academic researchers without overt褒
