What the EU's New AI Transparency Rules Mean for Your Content

Starting 2 August 2026, new EU rules require certain AI-generated content to carry clear disclosure. Here's a plain-language breakdown of what the regulation covers, when it applies, and what it means for the content you publish.

What the rule is

The obligations come from Article 50 of the EU AI Act, which addresses risks of deception and manipulation and aims to foster the integrity of the information ecosystem. These transparency obligations complement other AI Act rules, such as those for high-risk AI systems.

The rules cover two main areas:

Marking and detection of AI-generated content — a technical requirement placed on providers (the companies that build generative AI systems).
Labelling of deepfakes and certain AI-generated publications — a requirement placed on deployers (organizations and individuals who use generative AI systems to create content).

To help companies comply, the European Commission facilitated the creation of a Code of Practice on Transparency of AI-generated Content, developed by independent experts through a multi-stakeholder process. Adherence to the Code is voluntary, but the underlying transparency requirements in Article 50 are legal obligations regardless of whether an organization signs the Code.

The Commission and the AI Board have confirmed the Code is an adequate voluntary tool that organizations can use to demonstrate compliance with the AI Act's transparency obligations. By the end of July 2026, about 190 companies and organizations had signed it.

What the Code of Practice actually covers

The Code has two sections:

Section 1: Providers — rules for marking and detection of AI-generated and manipulated content.
Section 2: Deployers — rules for labelling of deepfakes and AI-generated and manipulated text.

The EU has also published a set of icons that deployers of generative AI systems may use to label AI-generated content.

The Code doesn't replace the AI Act itself or the Commission's official guidelines on Article 50 — it's a practical framework that signatories can rely on to demonstrate compliance. Organizations that choose a different compliance approach will need to demonstrate on their own that their measures are adequate; this gets assessed individually by national market surveillance authorities.

Who this affects

The Code was drafted with input from a broad group: providers of generative AI systems, developers of marking and detection technology, associations representing deployers, civil society organizations, academic experts, and organizations with expertise in transparency and very large online platforms.

Two working groups shaped the Code, mirroring the structure of Article 50:

Working Group 1 (Providers) focused on requiring that AI system outputs — audio, image, video, text — are marked in a machine-readable format and detectable as artificially generated or manipulated, using technical solutions that are effective, interoperable, robust and reliable as far as technically feasible.
Working Group 2 (Deployers) focused on requiring disclosure of content that is artificially generated or manipulated and constitutes a deepfake (image, audio or video resembling real people, objects, places, entities or events that could falsely appear authentic), as well as AI-generated or manipulated text publications on matters of public interest — unless that text has gone through human review and is subject to editorial responsibility.

Where things stand and how the Code came together

The Code was developed over roughly nine months, running from a September 2025 consultation through a closing plenary and publication of the final Code on 10 June 2026. Along the way it went through three drafting rounds, each followed by public drafts (published 17 December 2025 and 3 March 2026) and stakeholder workshops. Draft guidelines on the scope of Article 50 were published for consultation on 8 May 2026, and the Code is meant to work alongside those guidelines rather than duplicate them.