EU's AI Act Takes Effect, Forcing Transparency in AI-Generated Content

Published on 11/08/2026 18:38 in graphic-design • Tommy Wright

Index▼

EU's AI Act Takes Effect, Forcing Transparency in AI-Generated Content

The European Union's Artificial Intelligence Act, which came into force on August 2, is driving a significant change in the way AI-generated content is handled. In response to the new regulations, Anthropic, a leading AI company, has announced that it will start adding invisible watermarks to text produced by its latest AI models, including Claude.

These imperceptible watermarks, known as "machine-readable marking," will be woven into the generated text itself, making them difficult to remove. The watermarks won't affect the quality, readability, or meaning of the AI-generated content. They will, however, travel with the text when copied and pasted, potentially persisting through some editing.

The watermarking will be applied at the model level, meaning it will be present across all Claude products, surfaces, and platforms that utilize these AI tools. This includes Claude, the Claude API, Claude Code, Claude Cowork, and Claude Tag, as well as any platforms that integrate these models.

While the change is primarily driven by the EU's AI Act, Anthropic has stated that the invisible watermarking will be implemented globally, regardless of where the AI models are accessed. Models accessed through AWS, Google Cloud, or Microsoft Foundry will also feature the watermark.

The move towards transparency in AI-generated content is a significant step forward in addressing concerns around the potential misuse of AI. As AI technology continues to advance and become increasingly prevalent, such measures will likely become more widespread.

In the meantime, users can expect to see more AI-generated content, but with the added assurance that it has been marked as such. This development highlights the importance of responsible AI development and the need for clear guidelines around the use of AI-generated content.

What do you think about the introduction of invisible watermarks in AI-generated text? Let us know in the comments below.