Hisham Sarwar
The transparency obligations set out in Article 50 of the European Union’s AI Act became enforceable on August 2, 2026. From that date, providers and deployers of certain AI systems operating in the EU must tell people when they are interacting with an AI, label deepfakes and synthetic media, and disclose AI generated text published on matters of public interest. Non compliance can trigger fines of up to 15 million euros or 3 percent of worldwide annual turnover, whichever is higher.
What Article 50 actually requires
The European Commission published draft guidelines in May 2026 setting out how the rules are meant to work in practice. According to the Commission’s official guidance, the obligations fall into four groups.
First, systems designed to interact directly with people, including chatbots, virtual assistants and automated phone systems, must be built so that users are informed they are talking to an AI. Second, providers of systems that generate or manipulate synthetic content must mark that content in a machine readable way. Third, deployers of emotion recognition and biometric categorisation systems must inform the people exposed to them. Fourth, deployers of deepfakes and of AI generated text on matters of public interest must disclose that origin.
One detail matters more than the rest for anyone producing content. As legal analysis from Greenberg Traurig notes, a hidden machine readable watermark added by the model provider does not satisfy the deployer’s obligation. The disclosure has to be visible or audible and understandable without a detection tool. The rule also applies whether or not there was any intent to deceive.
A moving deadline
There is a complication worth flagging. The Council has announced a revised deadline of December 2, 2026 for these transparency obligations as part of a broader simplification package, but formal adoption of that change was still pending at the time of writing. The rules as currently in force apply from August 2, 2026. Anyone relying on the later date should confirm the current legal position rather than assume the extension has landed.
Why this is the important AI story right now
Most AI coverage tracks model releases and funding rounds. Regulation is less exciting and more consequential, because it changes what everyone building on top of these models is allowed to ship. The AI Act is the first comprehensive AI law from a major economy, and history suggests other jurisdictions will borrow from its structure the way many borrowed from GDPR.
The extraterritorial reach is the part people underestimate. The obligations attach to AI systems placed on the EU market or whose output is used in the EU. A studio in Lahore producing AI generated video for a client in Berlin is inside the scope of the rules, not outside them.
Practical steps for freelancers and small agencies
Start with an inventory. List every place your work touches generative AI: images, voiceovers, video, chat widgets on client sites, AI written copy that gets published. For each one, note who the provider is and who the deployer is, because the obligations differ.
Then decide on your disclosure language and apply it consistently. A short visible credit line on synthetic media, and a clear statement in a chat interface that the user is talking to an automated assistant, cover most everyday cases. Put the same terms into client contracts so responsibility is explicit rather than assumed.
Finally, keep records. Regulators generally look for a documented, good faith process before they look for perfection. Being able to show what you disclosed and why is worth more than a policy document nobody follows.
