BRUSSELS, BELGIUM / RankWire.AI / – European Union rules aimed at increasing transparency in artificial intelligence officially came into force across the bloc on August 2, 2026. Under Article 50 of the EU AI Act, the scope now includes chatbots, deepfakes, synthetic media, and some publicly relevant texts, with a clear distinction between technical marking obligations and notices visible to the audience. The legislation does not mandate that every AI-created item bears a visible label; instead, each requirement depends on factors such as the system used, the type of content, and how an organization presents it.
Organizations providing interactive AI systems are required to disclose when they are engaging with users through artificial intelligence, although they may refrain from doing so if a user can unmistakably recognize the AI nature of the interaction. Additionally, generative AI providers are obligated to incorporate machine-readable signals into synthetic text, audio, images, and videos, assisting detection tools in identifying generated or manipulated content. These signals should be applied with effective marking techniques whenever such methods are technologically feasible and dependable.
Organizations that publish synthetic media are subject to separate disclosure responsibilities, which include the obligation to identify deepfake images, recordings, and videos that might appear authentic, and, in certain cases, to label AI-generated texts related to public interest topics. However, this particular requirement can be bypassed if the content undergoes human review and editorial oversight, provided that the responsible individual or organization accepts accountability for the published material prior to the exemption being granted.
Visible Labels Focus on Deceptive Synthetic Content
The law also addresses systems that use emotion recognition or biometric categorization to analyze individuals, requiring operators to inform affected persons when these tools are employed, except in limited legal circumstances. Artistic, fictional, creative, and satirical works are given more flexible disclosure rules; notices should not disrupt the normal viewing or enjoyment of these works, yet organizations are still expected to provide an appropriate disclosure if such content contains deepfake elements.
The European Commission has issued guidance clarifying the transparency obligations, definitions, and available exemptions, while also introducing a voluntary code for developers and organizations deploying generative AI. Companies participating in the code can demonstrate their compliance with marking and labeling standards, whereas those outside the scheme must implement other effective measures to ensure adherence. While optional icons can help raise public awareness, they alone do not constitute proof of compliance under the EU AI Act.
Regulatory Bodies Are Empowered to Enforce the Rules
Enforcement primarily falls to national market surveillance authorities within each EU member state, with the European Data Protection Supervisor overseeing AI systems used by EU institutions and agencies. Violators may face fines of up to 15 million euros or 3% of their global annual turnover, whichever is lower; smaller enterprises will be subject to proportionate penalties based on the lesser of the two amounts. A transitional period is in place for generative AI systems introduced prior to August 2, 2026, granting providers until December 2, 2026, to comply with the machine-readable marking requirement. However, this extension applies solely to the technical marking obligation and does not delay the obligation to display notices for interactive AI, deepfake disclosures, or public-interest texts. Furthermore, the regulation does not require retroactive labels on content created before the transparency rules came into effect.