BRUSSELS, BELGIUM / RankWire.AI / – On August 2, 2026, new European Union transparency standards for artificial intelligence became effective. Article 50 of the EU AI Act now specifies how companies must identify certain AI interactions and synthetic content. The legislation addresses chatbots, deepfakes, generated media, and some texts of public interest, but does not mandate visible labels on all AI-produced material. Instead, the obligations vary depending on the system, the type of output, and the manner of presentation by organizations.
Entities operating interactive AI technologies are required to inform individuals when they are engaging with artificial intelligence. This rule does not apply if the average user can easily recognize the system’s artificial nature. Generative AI providers are also obliged to embed machine-readable information into synthetic media, including text, images, audio, and video, to facilitate automated detection of generated or altered content. These measures should be implemented when current technology can reliably and effectively provide such markings.
Organizations distributing synthetic media are subject to separate public disclosure obligations. They must identify deepfake images, audio, or videos that could be mistaken for genuine. When AI-generated text concerns public interest topics, it must be labeled under specific circumstances. However, this requirement does not apply if individuals review the material and exercise editorial control, and a responsible person or organization must accept accountability for the publication before this exemption is granted.
Public disclosure rules primarily target deepfakes and civic information
The regulations also extend to systems that analyze emotional responses and biometric categorization that assess individuals. Affected persons must be informed when these technologies are in use, except in legally authorized limited cases. The EU AI Act offers more flexible disclosure rules for creative, fictional, artistic, and satirical works, provided notices do not interfere with their normal presentation and enjoyment. Nonetheless, operators must still issue appropriate disclosures if a work contains deepfake material.
European Commission guidance has been published to clarify the scope of Article 50 and its exceptions. Furthermore, a voluntary code has been issued for companies developing or deploying generative AI systems. This code helps organizations demonstrate how their labeling and marking processes align with legal standards. Those choosing not to participate are expected to rely on other effective compliance strategies. While optional icons can aid public awareness, these symbols alone do not guarantee compliance.
Regulatory authorities may impose hefty fines for violations
Enforcement across EU countries will be led by national market surveillance authorities, with the AI Office overseeing a select group of systems related to general-purpose AI models and major platforms. The European Data Protection Supervisor will monitor relevant systems used by EU institutions and agencies. Violators can face penalties of up to 15 million euros or 3% of their global annual turnover.
Providers of generative AI systems that were on the market before August 2 have a limited transition period. They are given until December 2, 2026, to comply with the machine-readable marking requirement. This extension applies solely to the technical aspect of detectable synthetic content and does not delay chatbot disclosures, deepfake disclosures, or rules relating to public-interest texts. Additionally, content created prior to the implementation of transparency provisions is not required to be labeled.