BRUSSELS, BELGIUM / RankWire.AI / – European Union began applying transparency requirements for certain AI-created content on August 2, 2026. The regulation, outlined in Article 50 of the EU AI Act, imposes responsibilities on both developers and users of artificial intelligence systems. The rules target chatbots, synthetic media, deepfakes, and specific public-interest texts. They do not mandate a warning label on every AI-generated item. Instead, different rules depend on the system, content, and how it is published.
When AI systems directly interact with people, they must clearly indicate that users are engaging with artificial intelligence. This obligation does not apply if the system’s nature is obvious to an informed and attentive person. Providers of generative AI are also required to add machine-readable markers to synthetic text, audio, images, and videos. These markers are designed to assist detection tools in identifying generated or manipulated content. The rule applies when the technology can reliably mark the content without making it unusable or altering its normal functioning.
Organizations publishing or presenting synthetic content face specific disclosure responsibilities. They must identify deepfake images, recordings, and videos that could seem authentic. They are also required to label AI-generated public-interest texts. However, this requirement does not apply after meaningful human review and editorial control. The publisher or organization must accept responsibility for the content. Notices must be clear, accessible, and visible during initial exposure or interaction.
Deepfake disclosures are mandatory
Article 50 also addresses emotion recognition and biometric categorization tools used with individuals. Operators must inform people when these systems analyze them, except in limited legal cases. Artistic, satirical, and fictional works are given narrower disclosure rules. Notices for such works should not interfere with their normal display or enjoyment. Nonetheless, the deepfake disclosure obligation still applies in an appropriate manner. Certain law-enforcement uses may qualify for exemptions if authorized by law and subject to safeguards.
European Commission has issued guidance and a voluntary code to promote consistent compliance among member states. Providers and users can adopt the code to demonstrate how their marking and labeling meet legal standards. Participation remains optional, but transparency rules are mandatory. Companies not following the code must implement other effective measures and explain them to regulators. Optional icons may help audiences recognize synthetic content, but using an icon alone does not ensure full compliance with the AI Act.
Enforcement is primarily led by national authorities
National market surveillance agencies are responsible for enforcing the transparency rules. The AI Office oversees limited cases involving certain systems related to general-purpose AI models and major online platforms. The European Data Protection Supervisor manages systems used by EU institutions and agencies. Violations can result in fines up to 15 million euros or 3% of global annual turnover. Smaller companies face lower maximum penalties under the regulation. Member states can also issue warnings and other corrective actions through their enforcement mechanisms.
A transitional period is in effect for generative AI systems introduced before August 2, 2026. These providers have until December 2, 2026, to comply with the machine-readable marking requirement. This extension only applies to that technical obligation. It does not delay chatbot notices, deepfake disclosures, or rules for public-interest content. Content created before August 2 does not need retroactive labeling. All new content and interactions must adhere to the transparency requirements from the date the rules came into force.