AI Marker for your Website.

Brief Info

From a legal and design perspective, the EU AI Act is not just a pure compliance issue for me, but above all a question of transparency, trust, and user guidance. Regulation (EU) 2024/1689 has been in force since August 1, 2024, and is being applied gradually; for AI-generated content, Article 50 is particularly relevant because it describes transparency obligations for certain AI systems and their operators. Simply put, it's about people being able to recognize when they are interacting with an AI system or when texts, images, audio, or video content has been artificially generated or manipulated.

EU AI Act

The AI Act operates on a risk-based approach and differentiates between prohibited practices, high-risk systems, systems with transparency obligations, and general AI models. For websites, marketing content, editorial content, and visual works, it is particularly interesting that AI-generated or AI-manipulated content should not simply appear silently as "normal" content. In the case of deepfakes and AI-generated texts published to inform the public about matters of public interest, Article 50 provides for disclosure. The transparency obligations around Article 50 will become applicable from August 2, 2026, while the EU Commission is also working on guidelines and a Code of Practice to help providers and operators better assess what labeling should look like in practice.

AI Labeling

Labeling AI content is therefore not just about a small badge at the edge of an image. A sensible approach combines visible, understandable labeling for users and – where technically possible – machine-readable information, for example, through metadata, Content Credentials, or similar provenance approaches. Standards like C2PA show how the origin and editing of digital content can be made more traceable. Nevertheless, one should not blindly rely on metadata, as it can be lost depending on the platform, export, or further processing. A clearly visible notice therefore remains, in my view, the most pragmatic starting point.

Practical Implementation

My AI Marker Tool is designed precisely for this use case: it does not automatically solve every legal obligation and does not replace legal review, but it creates a clean, easy-to-understand basis for websites where AI-generated or AI-assisted content should not be hidden, but deliberately and transparently marked. For me, this is also a design question: the labeling should be visible without overlaying the content; clear enough to be understood, but subtle enough not to appear like a warning. This creates a transparent approach to AI that respects users, makes editorial work more comprehensible, and prepares early for future requirements.

Note: This classification is not legal advice, but a design and content orientation based on publicly available sources. For binding assessments, the specific application should always be legally reviewed.