en.Wedoany.com Reported - According to the European Commission, starting August 2, 2026, Article 50 of the European Artificial Intelligence Act, concerning transparency requirements, will officially come into effect. The European Commission has published official guidelines requiring providers and professional users (deployers) of AI systems to mark and label AI-generated or manipulated content, enabling citizens to identify its artificial origin.
The guidelines clarify the specific implementation of one of the most closely watched provisions of the EU AI Act—Article 50. Henna Virkkunen, Executive Vice President of the European Commission for Tech Sovereignty, Security, and Democracy, stated that the guidelines support the smooth and effective implementation of the Act, making AI systems that interact with humans (such as chatbots and agents) and automatically generated content more transparent and trustworthy, while helping citizens understand when they are interacting with content generated by this technology. The goal is to reduce the risks of deception, disinformation, and manipulation in a context where synthetic content is increasingly realistic and difficult to distinguish from authentic content, ensuring citizens can identify when they are interacting with an AI system or are exposed to AI-generated text, images, audio, or video.
The guidelines detail the obligations of AI system providers versus deployers (i.e., enterprises, public institutions, and professionals using these systems). Providers must design systems to clearly inform users when they are directly interacting with AI (e.g., chatbots) and integrate machine-readable markings into synthetic content, enabling other platforms and applications to identify its AI origin. Deployers must fulfill notification obligations to users when disseminating deepfakes, publishing AI-generated content of public interest without editorial control or human review, and when using emotion recognition or biometric categorization systems.
Markings are not only warnings visible to users but also technical metadata embedded in the content, allowing computer systems (such as digital platforms, search engines, verification tools, and regulatory bodies) to distinguish authentic content from synthetic content over the long term. The EU views Article 50 as a strategic pillar for building trust in AI applications, using transparency as a tool for consumer protection and information security.
The guidelines also specify exceptions: standard editorial operations, such as spelling or grammar corrections, do not transform content into AI-generated works and therefore do not require labeling. To facilitate compliance, the European Commission has also published a Code of Good Practices on Transparency of AI-Generated Content, developed by independent experts with the assistance of hundreds of stakeholders. This code is a voluntary tool that can be used to demonstrate compliance with the obligations set out in the AI Act.
Most obligations under Article 50 take effect on August 2, 2026, at which point the European Commission and national regulatory authorities can exercise the supervisory powers granted by the regulation. For AI systems already placed on the market before August 2026, a transitional period applies for obligations concerning the labeling and detection of synthetic content, which will come into effect from December 2, 2026. Additionally, the EU will establish an AI Act Service Desk as an information reference point to help businesses and organizations interpret the new regulations.
