BRUSSELS, BELGIUM / RankWire.AI / – European Union transparency standards for content generated by AI took effect throughout the bloc on August 2, 2026. Under Article 50 of the EU AI Act, entities are now required to provide explicit notices for certain AI interactions and synthetic media. The legislation applies to both providers responsible for creating AI systems and deployers who utilize these systems to publish or showcase content. Notably, there is no requirement to affix a visible label to every piece produced with AI. These obligations are part of the phased rollout following the law’s approval in 2024.
Entities offering systems that interact directly with individuals must inform users when they are engaging with AI. Such notices are not necessary if the artificial origin of the interaction is obvious to a reasonably informed person. Providers of systems that generate synthetic text, audio, images, or video are also required to include machine-readable indicators. These marks should enable detection of AI-created or manipulated content whenever technically feasible. Basic editing tools that do not significantly alter the input or its meaning are exempt from this requirement.
The law distinguishes between these technical marking responsibilities and the visible disclosures that deployers must provide to their audiences. Deployers must identify deepfake images, audio, and video when the material could falsely seem authentic. They are also tasked with labeling AI-generated texts published for the public, especially on issues of public interest. However, this text labeling obligation does not apply after a human review or editorial oversight, provided the responsible individual or organization accepts editorial accountability. All required disclosures must be presented clearly, in an understandable and accessible manner.
Distinctive machine markings are separate from public-facing labels
Article 50 extends its scope to encompass emotion recognition and biometric categorization systems, which mandate notices to those exposed. These disclosures should be presented clearly and distinctly by the first interaction or exposure. The law offers some exemptions for clearly artistic, creative, satirical, or fictional works to preserve artistic display and audience enjoyment. For deepfakes, disclosures can be minimized within creative works but cannot be omitted entirely. Additionally, specific exceptions are provided for certain law enforcement uses authorized by law and subject to safeguards.
On July 20, the European Commission issued comprehensive guidelines clarifying the scope, definitions, and exceptions related to these rules. It also released a voluntary Code of Practice on June 10 aimed at providers and deployers of generative AI systems. Companies can adopt this code to demonstrate their compliance with labeling and marking responsibilities across the EU. Organizations that do not sign up must implement other suitable measures and prove to authorities that these measures meet legal requirements. The Commission additionally introduced optional icons, but these alone do not constitute legal compliance.
Primary enforcement is handled by national authorities
Responsibility for enforcing the new transparency rules primarily rests with national market surveillance authorities. The AI Office manages limited cases involving systems linked to general-purpose AI models or large online platforms. The European Data Protection Supervisor oversees systems employed by EU institutions, bodies, and agencies. Penalties for violations can reach up to 15 million euros or 3% of global annual turnover, with smaller businesses facing proportionate limits. Penalty procedures are established by member states, which may use warnings or other non-monetary enforcement actions.
A transitional period applies for generative AI systems introduced before August 2, 2026. Providers of these existing systems have until December 2, 2026, to meet the machine-readable marking requirement. This transition only pertains to the technical marking and detection obligations outlined in Article 50 and does not delay other transparency duties, which are already in effect across the European Union. Content produced before August 2 does not need retroactive labeling, while new material must adhere to the relevant disclosure and marking standards.