BRUSSELS, BELGIUM / RankWire.AI / – European Union transparency regulations concerning artificial intelligence became effective across the bloc on August 2, 2026. Article 50 of the EU AI Act now encompasses chatbots, deepfakes, synthetic media, and certain public-interest texts, establishing a distinction between technical marking responsibilities and notices visible to audiences. The legislation does not stipulate that every AI-created item must bear a visible label, but instead assigns obligations based on the system, the content, and how an organisation presents it.
Entities providing interactive AI systems are required to notify users when they are engaging with artificial intelligence, although this can be omitted if a knowledgeable user can easily recognize the system’s artificial origin. Generative AI vendors are also mandated to incorporate machine-readable signals into synthetic texts, audio, images, and videos, aiding detection tools in identifying manipulated or generated material. These signals should be implemented with effective marking methods when current technology makes them feasible and dependable.
Organizations distributing synthetic content are subject to distinct disclosure obligations, including the need to identify deepfake images, recordings, and videos that could appear authentic. They must also label AI-generated texts related to public interest issues in specific cases, although human oversight and editorial review can exempt certain requirements. Responsibility for the published material must be accepted by an individual or organisation prior to claiming any exemptions.
Visible notices aim at deceptive synthetic media
Article 50 also regulates the use of emotion recognition and biometric categorisation systems that evaluate individuals, requiring operators to inform those affected, barring limited legal exceptions. The law allows more flexibility for artistic, fictional, creative, and satirical works, with notices that should not disrupt their typical presentation or enjoyment. Nonetheless, organisations must provide appropriate disclosures when such media contain deepfake content.
The European Commission has published guidance detailing the transparency obligations, definitions, and exceptions available. It also introduced a voluntary code for developers and organisations implementing generative AI, enabling participants to demonstrate compliance through their marking and labelling strategies. Companies not adhering to the code must adopt other effective measures, and while optional icons may raise public awareness, they do not alone verify compliance with the EU AI Act.
Authorities are empowered to enforce rules
Market surveillance authorities within each EU member state hold primary enforcement responsibilities, with the European Data Protection Supervisor overseeing AI systems used by EU institutions and agencies. Violations of transparency requirements can lead to fines of up to 15 million euros, or 3% of a company’s global annual turnover, whichever is lower. Smaller businesses are subject to reduced maximum penalties based on this calculation.
For generative AI systems introduced before August 2, 2026, a transitional period is in place, giving providers until December 2, 2026, to comply with the machine-readable marking obligation. This extension solely applies to the technical marking requirement and does not delay disclosures related to interactive AI, deepfake notices, or public-interest texts, nor does it require retroactive labels for content created prior to the law’s enforcement date.