EU AI Act Article 50 Explained: Deadlines, Scope, and Website Impact

If you’re trying to understand what Article 50 of the EU AI Act requires of your website, this guide breaks down each obligation, who it applies to, what content it covers, and the deadlines you need to know. For the complete compliance overview, see our EU AI Act compliance guide for WordPress.

What Is Article 50?

Article 50 of the EU AI Act (Regulation (EU) 2024/1689) sets out transparency obligations for providers and deployers of AI systems. These obligations are designed to address risks of deception and manipulation by ensuring people know when they’re interacting with AI and when content has been artificially generated or manipulated (European Commission).

The obligations are not about banning AI content — they’re about transparency. You can publish AI-generated content; you just need to tell people it’s AI-generated in the ways Article 50 specifies.

The Five Obligations at a Glance

Source: European Commission FAQ

Who Must Comply?

Providers

Providers are entities that develop AI systems or place them on the EU market under their own name or trademark. This includes companies like OpenAI, Google, and Anthropic, but also any organization that develops and deploys its own AI system. Providers must comply with their obligations before placing systems on the market or into service, regardless of where they are established — if the AI system’s output is used in the EU, the AI Act applies.

Deployers

Deployers are entities that use AI systems under their authority in a professional context. Key points:

• Personal, non-professional use is excluded
• If the activity generates economic benefit on a regular basis or is connected with a business, trade, or freelance activity, it’s professional activity
• Employees acting under an organization’s instructions are not separate deployers — the organization is
• A legal person remains the deployer even when contractors or freelancers operate the system on its behalf

For most WordPress site owners, the relevant question is: “Does my organization use AI in a professional context?”

If yes — and the output is used in the EU — Article 50 likely applies.

Article 50(1): AI Chatbot Disclosure

If your website runs an AI chatbot, AI agent, or avatar that directly interacts with natural persons, the provider of that system must ensure people are informed they are interacting with AI. The obligation applies cumulatively where:

1. The system qualifies as an AI system
2. It’s designed for genuine two-way exchange with people
3. The interaction is direct (the AI communicates with the person, not through a human intermediary)
4. The interaction is with natural persons (consumers, professionals, or other users)

Notification must be given from the start of the first interaction and must be clear, distinguishable, and accessible. The “obvious” exception — where it’s apparent to a reasonably informed person that they’re interacting with AI — is interpreted restrictively (European Commission).

For practical implementation on WordPress, see our AI chatbot disclosure requirements guide.

Article 50(2): Machine-Readable Content Marking

Providers of generative AI systems must ensure that AI-generated or manipulated audio, images, video, and text are marked in a machine-readable format. The marks must be:

• Effective
• Reliable
• Robust
• Interoperable

This is a provider-side obligation — but deployers should not strip these marks from content they publish. Exemptions exist for standard editing assistance, short character sequences, source code, and machine-to-machine outputs.

The Code of Practice on Transparency of AI-generated Content, published 10 June 2026, provides a voluntary framework for providers to demonstrate compliance. Non-signatories must demonstrate compliance through alternative adequate means.

Article 50(4): Deepfake and AI Text Labeling

This is the obligation most directly relevant to WordPress publishers. Deployers of generative AI systems must clearly label:

Deepfakes — AI-generated or manipulated image, audio, or video content that:

1. Resembles existing persons, objects, places, entities, or events
2. The subject exists or could plausibly exist
3. Would falsely appear authentic or truthful to a person

All three criteria must be met cumulatively. Deployers must disclose deepfakes no later than a person’s first exposure to the content. Disclosure must be clear, distinguishable, understandable, and perceivable — visible labels or audible labels are suitable. Deployers cannot rely solely on the machine-readable marks embedded by the provider (European Commission).

For artistic, creative, satirical, or fictional works, the transparency obligation is lighter — disclosure must not hamper display or enjoyment of the work. See our deepfake disclosure rules guide for full details.

AI-generated text on matters of public interest must be labeled unless it has undergone genuine human review or editorial control. Three cumulative criteria determine whether text falls under this obligation:

1. The text must be published
2. The text must be informative to the public
3. The text must concern matters of public interest — including politics, public administration, justice, fundamental rights, public security, public health, environmental protection, consumer safety, and economic, financial, political, scientific, or cultural developments relevant to public debate

The exemption requires genuine human review: deliberate examination of the substance by qualified persons with relevant knowledge and professional judgment, conducted under editorial responsibility with authority to approve, alter, or reject the content. Spell-checking, grammar correction, and superficial or procedural checks do not qualify (European Commission).

For practical labeling guidance, see how to label AI-generated content.

Deadlines

Source: European Commission

Enforcement and Fines

• Enforced by national competent market surveillance authorities
• The AI Office has a limited role for GPAI-integrated systems and very large online platforms/search engines
• Fines: up to €15 million or 3% of total worldwide annual turnover (whichever is higher)
• Proportionality considerations for SMEs and small mid-caps

What This Means for Your WordPress Site

If your site uses AI chatbots, publishes AI-generated text on matters of public interest, or hosts AI-generated images, audio, or video, you need to:

1. Disclose AI chatbots from the first interaction
2. Label AI-generated text that hasn’t undergone genuine human review
3. Label deepfakes clearly and visibly
4. Preserve machine-readable marks on AI-generated content
5. Keep records of your compliance efforts

Tools like MegaGovern can help WordPress site owners implement these practical measures — from disclosure labels to evidence preservation. For a full step-by-step approach, see our compliance checklist, and for a comparison of available tools, see our best AI disclosure plugins guide.

Disclaimer: This article is informational, not legal advice. Consult the official European Commission guidance and a qualified legal professional for your specific situation.