European Commission: AI Transparency Guidelines Under Article 50 Take Effect August 2

What does Article 50 of the EU AI Act require?

Article 50 is part of the EU AI Act (the EU’s Artificial Intelligence Regulation), the European law governing the development and use of AI systems across the single market. The provision introduces transparency obligations for stakeholders that develop or use AI technology.

The European Commission has published guidelines for complying with these obligations. The guidelines take effect on August 2, 2026, together with an FAQ document and an accompanying Code of Practice for the transparency of AI-generated content.

The EU AI Act follows a risk-based approach under which stricter rules apply to systems with greater potential for harm, and transparency is one of the foundational pillars of the entire regulation.

Provider obligations: design and machine-readable markers

Providers are companies that develop and place AI systems on the market. Under the new guidelines, they must design systems so that users are clearly informed when they are interacting directly with artificial intelligence rather than a human.

An additional obligation is embedding machine-readable markers in any content that AI has generated or modified. The markers must be recognizable by automated tools, not just to the human eye.

Such markers allow platforms, search engines, and other tools to automatically identify AI content, instead of relying solely on human judgment.

Deployer obligations: deepfakes and biometrics

Deployers are organizations that use other providers’ AI systems in practice, such as media outlets or public institutions. They must inform people when they are exposed to one of three categories of content.

The first is a deepfake — an image, video, or audio recording that falsely depicts a person in a situation that did not occur. The second is AI-generated content on matters of public interest published without editorial oversight. The third is emotion-recognition or biometric-categorization systems.

Unlike providers, whose obligation is built into the product’s design itself, deployers are responsible for informing end users at the point of use — two layers of accountability covering the entire chain from production to the application of AI content.

Why is the August 2 deadline critical for every AI company in the EU?

The deadline matters because the obligations apply to every company that places or uses AI systems on the EU market, regardless of where it is headquartered. That includes major global players such as OpenAI, Google, and Anthropic, as well as local startups.

Non-compliance after August 2, 2026, exposes companies to sanctions under the broader EU AI Act framework. The Code of Practice also offers concrete technical examples of labeling, narrowing the room for divergent interpretations of the obligation across the 27 member states.

The obligation applies regardless of whether a company offers its services directly from the EU or reaches European users over the internet from a third country — a common extraterritorial reach of European digital regulation.