BRUSSELS, BELGIUM / RankWire.AI / – Effective from August 2, 2026, new European Union transparency regulations for artificial intelligence have come into force. Under Article 50 of the EU AI Act, companies are now required to specify certain AI interactions and synthetic content. The legislation addresses chatbots, deepfakes, generated media, and some public-interest text, but it does not mandate visible labels on all AI-produced material. Instead, the obligations depend on the specific system, output, and how an organization displays it.
Operators of interactive AI systems must inform users when they are engaging with artificial intelligence. This requirement does not apply when an average user can easily recognize the AI’s artificial nature. Generative AI providers are also required to embed machine-readable information into synthetic text, images, audio, and video to facilitate automated detection of modified or generated content. These measures should be implemented when existing technology can reliably and effectively support marking.
Entities distributing synthetic media face distinct public disclosure rules. They must identify deepfake images, audio, or video that could appear authentic and label AI-generated content related to public interest topics in specific cases. This labeling obligation does not apply if individuals review the material and exercise editorial control. Additionally, a named person or organization must accept responsibility for the publication before this exemption can be claimed.
Public disclosure requirements primarily target deepfakes and civic information
The regulations extend to systems that evaluate emotions and biometric data, which assess individuals. Operators must notify affected persons when employing such technologies unless legally authorized exceptions apply. The EU AI Act affords more flexible disclosure rules for creative, fictional, artistic, and satirical works, as long as notices do not disrupt their presentation or enjoyment. Still, operators must disclose the presence of deepfake material when relevant.
European Commission guidance clarifies the scope of Article 50 and its exemptions, alongside a voluntary code for developers and users of generative AI systems. This code allows companies to demonstrate how their labeling and marking methods comply with legal standards. Non-participating companies must rely on other effective compliance measures, with optional icons available to enhance public awareness, although these symbols alone do not ensure compliance.
Regulatory bodies can levy substantial fines for violations
Enforcement will primarily be carried out by national market surveillance authorities across EU member states, while the European Commission AI Office will oversee a smaller subset of systems linked to general-purpose AI models and major platforms. The European Data Protection Supervisor will supervise relevant 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.
Systems of generative AI already available before August 2 have a transitional period until December 2, 2026, to comply with the machine-readable marking requirement. This extension only applies to detectable synthetic content; it does not delay chatbot notices, deepfake disclosures, or rules related to public-interest content. The law also clarifies that organizations are not obliged to label content created before the transparency rules took effect.