EU AI Act Chatbot Rules: What an Online Store Has to Tell Its Customers
The EU AI Act has chatbot rules, and they're worth reading if your store has an AI chat and sells to people in the EU. For a support chatbot, the part that matters is one article, and it's short. This post explains the rule. It isn't legal advice.
What does the AI Act say about chatbots?#
Article 50 is the AI Act's transparency article. Its first paragraph is the one that covers a support chat:
"Providers shall ensure that AI systems intended to interact directly with natural persons are designed and developed in such a way that the natural persons concerned are informed that they are interacting with an AI system, unless this is obvious from the point of view of a natural person who is reasonably well-informed, observant and circumspect, taking into account the circumstances and the context of use."
Article 50(1), AI Act Service Desk, read 15 October 2026
In plain words: if a customer could mistake your chat for a person, it has to tell them it's an AI.
When did the rule start?#
On 2 August 2026. Article 113, which sets the AI Act's dates, says it "shall apply from 2 August 2026" (Article 113, read 15 October 2026). The AI Act Service Desk's implementation timeline lists the same day, with the line "Transparency rules (Article 50) start to apply". The European Commission's guidelines page on transparency obligations says Article 50 applies from 2 August 2026 as well.
Who has to do it: the store or the software?#
Article 50(1) puts the design duty on the provider, the company that builds the AI system. The Commission's guidelines describe that duty as making sure people are "explicitly informed whenever they interact with an AI system directly". If you run a vendor's chatbot on your store, you're usually the deployer rather than the provider.
You still have your part to play, mind. Pick a tool that labels its replies as AI. Don't switch the label off where a tool lets you, and don't write instructions telling the bot to pass as a person. A label you've hidden doesn't inform anyone.
What should the disclosure look like?#
Article 50(5) says the information has to be given "in a clear and distinguishable manner at the latest at the time of the first interaction or exposure", and that it "shall conform to the applicable accessibility requirements".
In a store chat, that usually comes down to one of two things:
- a label on each AI reply, visible from the first message
- a first message that says plainly it's an AI assistant
Neither example is a legal standard. They're the obvious ways to meet the wording, and your adviser can tell you whether your setup does. Accessibility counts too, so a text label a screen reader can read out beats an icon on its own.
Does this change GDPR's right to a person?#
No. Article 50 sits alongside GDPR. Article 22 of GDPR gives people the right not to be subject to a decision based solely on automated processing that has legal or similarly significant effects on them. Where such a decision is allowed, Ireland's Data Protection Commission lists the safeguards: the right to obtain human intervention, to present your point of view and to challenge the decision. Telling a customer they're talking to an AI doesn't replace any of that.
How Hay.chat handles it#
Replies carry an AI label by default, and EU workspaces can't hide it. The agent's own instructions tell it never to pretend to be a person. When a customer asks for one, the conversation goes to your team.
The label only covers telling people it's AI. For the answers themselves, see how Hay.chat checks them before they go out.
Frequently asked questions
Hay.chat labels replies as AI by default, and EU workspaces can't hide the label. See how it looks to your customers in the 30-day free trial, no credit card.
Sources#
All pages read 15 October 2026.
- AI Act Service Desk (European Commission): Article 50, transparency obligations
- AI Act Service Desk: Article 113, entry into force and application
- AI Act Service Desk: implementation timeline
- European Commission: guidelines on AI transparency obligations, page updated 6 August 2026
- Data Protection Commission (Ireland): your rights in relation to automated decision-making
Ask AI about this page
About the Author
Damien Mulhall
Co-Founder, Strategy and Content
Damien spent 10+ years managing support operations and project delivery for global brands including Dell, Microsoft, Intel, and Google. He's PMP-certified, and that operations background shapes how Hay.chat is built.
