BRUSSELS, BELGIUM / RankWire.AI / – Effective from August 2, 2026, the European Union has implemented new transparency standards for artificial intelligence. As outlined in Article 50 of the EU AI Act, these rules specify how organizations must identify certain AI-driven interactions and synthetic material. The legislation encompasses chatbots, deepfakes, generated media, and some texts of public interest. However, it does not mandate visible labels for all AI-produced content. Instead, the requirements vary based on the system, the type of output, and how organizations present it.
Operators of interactive AI systems are obligated to inform users when they are engaging with artificial intelligence. This obligation does not extend to situations where an ordinary user can easily recognize the artificial nature of the system. Providers of generative AI are also required to embed machine-readable metadata within synthetic text, images, audio, and video. Such information facilitates automated detection of altered or generated content. The requirement applies when current technology can reliably and effectively implement such marking.
Distinct disclosure rules are set for entities distributing synthetic media. They must identify deepfake images, audio, or video that could be mistaken for real. Additionally, AI-generated texts about public-interest topics must be labeled under certain conditions. This requirement does not apply when individuals review the material and exercise editorial control. Furthermore, a specific person or organization must accept responsibility for the publication prior to this exemption taking effect.
Focus on public disclosures for deepfakes and civic information
The regulation also addresses emotion recognition and biometric categorization systems that analyze individuals. Operators are required to notify affected parties when employing such technologies, except in limited legally authorized cases. The European Commission has provided guidance clarifying the scope of Article 50 and its exceptions. A voluntary code has been issued for companies involved in developing or deploying generative AI. Organizations can use this code to demonstrate compliance with labeling and marking standards. Those who opt out will need to rely on alternative effective measures. While optional icons may promote public awareness, these symbols alone do not constitute compliance.
Enforcement measures include hefty fines for non-compliance
Enforcement will primarily be carried out by national market oversight authorities across EU member states. The European Data Protection Supervisor will supervise systems used by EU institutions and agencies, while the AI Office will focus on general-purpose AI models and major platforms. Penalties for violations can reach up to 15 million euros, or 3% of a company’s global annual turnover. The law provides a transitional period for AI systems already available before August 2. Their providers have until December 2, 2026, to comply with the machine-readable marking requirement for synthetic content detection. This extension does not affect disclosures related to chatbots, deepfakes, or public-interest texts, nor does it require labeling of content produced before the new transparency rules came into effect.