For months, the conversation around the EU's Artificial Intelligence Act has focused on the obligations of high-risk systems and the handful of tech giants that dominate the market. But as of 2 August 2026, a new set of transparency rules under Article 50 of the Act has widened the net considerably. Now, individual creators, freelancers, small agencies, and even social media influencers who produce or deploy AI-generated content for the EU market are subject to binding requirements—or risk penalties that can reach millions of euros.
The rules apply not only to entities based in the EU but also to those outside the bloc, provided the AI systems or content are used in the EU market. Whether the service is paid or free of charge makes no difference. This means a YouTuber in London, a podcast producer in Berlin, or a graphic designer in Belgrade—if their content reaches EU audiences—must now comply.
Four key obligations under Article 50
The European Commission has issued guidelines to clarify the four main transparency duties. First, AI systems that interact directly with people—such as chatbots, voice assistants, and AI avatars—must be designed so that users are clearly aware they are dealing with a machine. Second, AI systems that generate or manipulate synthetic content (images, video, audio, or text) must embed machine-readable markers to facilitate detection. Third, systems used for emotion recognition or biometric identification require deployers to inform individuals that they are being exposed to such technology. Fourth, any AI-generated or manipulated content that touches on “matters of public interest”—including health, public security, justice, or environmental protection—must be disclosed to the public.
These obligations are not limited to large corporations. The guidelines explicitly cover “individual and legal persons,” which includes media outlets, NGOs, advertising agencies, and even solo freelancers. The only exemptions are for purely personal use, research and scientific purposes, open-source systems, and artistic or satirical content—though the lines can be blurry, and the Commission advises taking the guidelines seriously.
Who is affected and what must be labelled?
For writers and bloggers, AI-generated text that substantially changes the style, structure, or meaning of the original must be marked. This includes paraphrasing that alters the message. However, minor grammar corrections, spellchecking, and stylistic polishing are exempt. For social media influencers, replacing or substantially modifying faces in photos—especially for commercial purposes—requires a label. Graphic and video designers must mark content where objects or people are removed, replaced, or inserted, or where body shape or skin colour is altered. Even extreme colour or contrast changes that shift the meaning of the content fall under the rules.
Podcasters and audio editors must label synthetic speech that mimics a specific person's voice. The same applies to videos depicting events that never occurred. The key principle is transparency: if AI has materially changed the content, audiences should know.
For those who rely on AI for minor edits—like cropping, red-eye removal, or video stabilisation—no labelling is required. But the Commission warns that “substantial change” is not always easy to define, and editorial review must go beyond simple proofreading. A named individual must take responsibility for the final content.
Fines and enforcement
Non-compliance can be costly. Providers and deployers face fines up to €15 million or 3% of their total worldwide annual turnover, whichever is higher. EU institutions themselves are not exempt and could be fined up to €750,000. The European Commission has said it will enforce the rules from the start, and national authorities across the 27 member states are expected to coordinate.
The new obligations come at a time when AI-generated content is increasingly common in everyday life. From deepfakes and chatbots to AI-assisted design, the EU is trying to ensure that citizens can distinguish between human and machine creation. The rules also build on earlier efforts, such as the Brussels finding that TikTok failed to protect minors' privacy, showing a broader regulatory push to hold digital actors accountable.
For businesses and creators, the message is clear: review your AI workflows now. The summer recess may be over, but the obligations are just beginning. As the Commission’s guidelines emphasise, transparency is not a burden but a way to build trust in the digital single market.
While some may see the rules as an administrative headache, they also level the playing field. Small creators and independent media now have clear standards, just like the tech giants. The EU is betting that transparency will foster innovation, not stifle it. Whether that bet pays off will depend on how smoothly the new rules are implemented across the continent.
