BRUSSELS, BELGIUM / RankWire.AI / – Effective from August 2, 2026, the European Union introduced new transparency regulations for artificial intelligence. Article 50 of the EU AI Act now outlines how organizations must identify certain AI-driven interactions and synthetic materials. The legislation addresses chatbots, deepfakes, AI-generated media, and specific public-interest texts. However, it does not mandate visible labels on all AI-produced content. Instead, obligations vary based on the system used, the output produced, and the manner of presentation by the organization.
Operators of interactive AI systems are required to notify users when they are engaging with artificial intelligence. This rule does not apply if the average user can easily recognize the artificial nature of the system. Generative AI providers are also mandated to incorporate machine-readable data into synthetic text, images, audio, and video content. This embedded information aims to facilitate automated detection of altered or AI-generated materials. The requirement applies when current technology can reliably deliver effective marking.
Organizations distributing synthetic media must adhere to separate disclosure rules for the public. They are responsible for identifying deepfake images, audio, or video that may seem authentic. Additionally, AI-generated public-interest texts must be labeled in particular cases. This labeling exemption does not apply if the material is reviewed and editorial oversight is exercised by a person or organization. The responsible party must also be identified and accept accountability for the published material before the exemption applies.
Disclosures focus on deepfakes and civic-related content
The regulations extend to systems that analyze individuals through emotion recognition or biometric categorization. Operators must inform affected parties when deploying these technologies, except in specific cases permitted by law. The EU AI Act offers more flexible disclosure options for creative, fictional, artistic, and satirical works. Such disclosures should not interfere with the normal viewing or enjoyment of the content. Nonetheless, operators are still obliged to disclose when a work contains deepfake material.
The European Commission issued guidance clarifying the scope of Article 50 and its exceptions. It also launched a voluntary code of conduct for firms developing or deploying generative AI systems. This code enables companies to demonstrate how their labeling and marking processes comply with legal standards. Organizations choosing not to participate must implement other effective compliance measures. Although optional icons can help raise public awareness, they do not, by themselves, confirm regulatory compliance.
Authorities may levy substantial financial penalties
National market surveillance agencies will oversee enforcement within EU member states. The AI Office will monitor a select group of systems linked to general-purpose AI models and major platforms. The European Data Protection Supervisor is tasked with supervising relevant systems used by EU institutions and agencies. Fines for non-compliance can reach up to 15 million euros, with penalties that can also be as high as 3% of a company’s global annual turnover.
Existing generative AI systems introduced before August 2 benefit from a transitional period. Their developers have until December 2, 2026, to comply with the machine-readable marking requirement. This extension only applies to the technical aspect of detecting synthetic content and does not postpone chatbot disclosures, deepfake notifications, or rules related to public-interest texts. Moreover, organizations are not obligated to label content created prior to the implementation of transparency rules.