BRUSSELS, BELGIUM / RankWire.AI / – Effective from August 2, 2026, new European Union transparency standards concerning artificial intelligence have been enforced. Under Article 50 of the EU AI Act, rules now specify how companies must identify certain AI-driven interactions and synthetic media. The regulation addresses chatbots, deepfakes, generated multimedia, and some publicly relevant texts. It does not mandate visible labels on all AI-produced content; instead, responsibilities vary depending on the system, output, and how an organization presents the material.
Operators of interactive AI systems are required to notify users when they are engaging with artificial intelligence. Exceptions are made when the system’s artificial nature is obvious to the average user. Providers of generative AI must also embed machine-readable data within synthetic images, audio, video, and text, facilitating automated identification of manipulated or generated content. The technology must be capable of reliably and effectively marking such content when available.
Organizations distributing synthetic media are subject to separate public disclosure mandates. They are required to label deepfake images, audio, or video that may appear authentic. Additionally, in certain circumstances, AI-generated texts addressing public-interest topics must be identified. However, these requirements are waived if individuals review the material and exercise editorial oversight. Responsibility for publication must be clearly attributed to a specific person or organization before this exemption applies.
Transparency obligations primarily target deepfakes and civic information
The regulations extend to emotion detection and biometric categorization tools used to evaluate individuals. Operators must inform those affected when deploying such technologies, except in legally authorized limited cases. The EU AI Act provides more flexible disclosure rules for creative, fictional, artistic, and satirical works, requiring notices that do not disrupt the work’s presentation or enjoyment. Nonetheless, operators must still include appropriate disclosures if a work contains deepfake elements.
The European Commission issued guidance clarifying the scope of Article 50 and its exceptions, along with a voluntary code for companies involved in developing or deploying generative AI. This code enables companies to demonstrate how their labeling and marking processes align with legal standards. Those opting out of the code must employ other effective compliance strategies. While optional icons can enhance public understanding, they do not, on their own, fulfill legal compliance requirements.
Regulatory bodies hold the authority to impose substantial fines
Supervisory authorities at the national level will oversee enforcement across EU nations. The AI Office will monitor a limited set of systems linked to general-purpose AI models and major platforms. The European Data Protection Supervisor is responsible for systems used by EU institutions and agencies. Penalties for violations can reach up to 15 million euros or 3% of a company’s global annual turnover. Enforcement powers include the ability to impose these fines for non-compliance.
Existing generative AI systems on the market before August 2 are granted a transitional period. Their providers have until December 2, 2026, to comply with the machine-readable marking requirement. This extension concerns solely the technical aspect of detectable synthetic content and does not affect chatbot disclosures, deepfake notices, or rules related to public-interest texts. Additionally, organizations are not obliged to label content created prior to the regulation’s enforcement date.