Chatbot Disclosure under the EU AI Act: What Every Website Owner Must Know

If your website has a chat widget, a virtual assistant, or any kind of automated conversational interface, you have a new legal obligation coming into force on 2 August 2026. Article 50(1) of the EU AI Act (Regulation 2024/1689) requires that users be clearly informed when they are interacting with an AI system — not a human. This applies to virtually every chatbot in use today.

This guide explains exactly what the rule requires, who falls under it, what a proper disclosure must say, and how to implement it on your website before the deadline.

What Article 50(1) Actually Says

Article 50(1) of the EU AI Act states that deployers of AI systems intended to interact with natural persons must ensure those persons are informed they are interacting with an AI system — in a clear and distinguishable manner — at the latest at the beginning of the interaction.

There are two key phrases here. "Clear and distinguishable" means the disclosure cannot be buried in footnotes, collapsed into small print, or camouflaged within the conversation. It must be prominent. "At the beginning of the interaction" means before or at the moment the user sends their first message — not after, not on a separate page, not in the Terms of Service.

The only exception is when the AI nature of the interaction would be obvious to any reasonable person in context. This exception is narrow and does not apply to general-purpose customer support or sales chatbots, which are specifically designed to feel conversational and human-like.

Who Does This Apply To?

The obligation falls on deployers — the businesses and individuals who put AI systems to use on their websites. If you have embedded any of the following on your site, you are a deployer under the Act:

It does not matter whether the chatbot vendor handles the underlying AI — if you have deployed the widget on your site and it interacts with your visitors, the obligation is yours as the deployer. Some vendors may provide compliance features to help (such as an automatic disclosure message), but you cannot simply assume they have done this without verifying.

The regulation applies if you target EU users — meaning your site is accessible in the EU, or your content, pricing, or language options suggest you serve EU customers. This includes businesses incorporated outside the EU.

What a Proper Disclosure Must Say and Where It Must Appear

The law does not mandate specific wording, but it does require the message to be clear and unambiguous. In practice, the following elements should be present:

Example compliant disclosure message:

"Hi there! I'm an AI assistant here to help with your questions. You're chatting with a bot, not a human agent. If you need to speak with a person, type 'human' or email us at support@example.com."

The disclosure must appear as the first visible message in the chat interface, displayed automatically when the widget opens — before the user types anything. It cannot appear only after the user sends a message, and it cannot rely solely on a small icon or label at the edge of the widget. The text itself must make the AI nature explicit.

Common Mistakes That Will Not Be Enough

Hiding disclosure in Terms and Conditions

Many website owners assume that a clause in their Terms of Service ("By using our chat function, you acknowledge it may be AI-powered") satisfies the requirement. It does not. Article 50 requires disclosure at the point of interaction, not in a document most users never read.

Showing disclosure only after the first message

Some chat platforms display a note such as "You are chatting with an AI" only after the user has already sent a question. The regulation is explicit: disclosure must occur "at the beginning of the interaction" — which regulators are likely to interpret as before the first user input.

Relying on an avatar or robot icon alone

Visual cues like a robot avatar or a small "AI" badge may support compliance, but they are unlikely to satisfy the "clear and distinguishable" standard on their own. Explicit text remains the safest approach.

Vendor-provided disclosures you have not verified

Check your chatbot's actual greeting message right now. Log into your vendor dashboard, find the welcome or first message setting, and verify that it clearly identifies the system as AI. Do not assume the vendor has done this for you.

How to Add the Disclosure on WordPress

For most WordPress chatbot plugins, the disclosure is added through the plugin's settings panel. Look for a "Welcome message," "Opening message," or "Bot greeting" field. Add language that makes the AI nature explicit as the very first thing a visitor reads.

If you are using a widget embedded via a JavaScript snippet (common with Intercom, Drift, or Zendesk), check the vendor's dashboard for conversation settings or automation rules. Most platforms have a "first message" or "proactive message" feature that fires as soon as the widget opens.

For custom-built chatbots or OpenAI API integrations, the disclosure should be hardcoded as the initial system-side message, rendered in the chat UI before the user can type.

Not sure if your current chatbot has a visible disclosure? Use legibright.eu to scan your site and get a compliance report in seconds. It checks for chatbot disclosure signals alongside other Article 50 requirements, so you can see exactly what is missing before August 2.

This article is for general information only and does not constitute legal advice. For your specific situation, consult a qualified legal professional.