Chatbot on your webshop? These transparency obligations apply now

Chatbots have become indispensable on webshops. They answer product questions, help customers with their orders, and take over part of customer service. Handy for you and for your customer.

But since 2 August 2026, specific transparency obligations have applied in Europe for AI systems that communicate directly with people. A chatbot may therefore not simply pass itself off as a human member of staff.

What does this mean in practice for your webshop?

Your customer must know they're talking to AI

The key rule is simple: when someone communicates directly with an AI system, that person must be informed of this.

Your customer must therefore be able to see clearly that they are talking to a chatbot or virtual assistant, and not to a human member of staff. Under Article 50 of the European AI Act, this information must be provided at the latest upon first contact.

A mention that only appears somewhere at the bottom of your terms and conditions or privacy policy is therefore not sufficient. The information must be visible from the moment the customer starts using the chatbot.

AI-generated content must be identifiable Does your chatbot also generate text, images or other material that is used elsewhere (for example, automatically generated product descriptions)? Where relevant, this must then be marked as AI-generated in a machine-readable format.

Emotion recognition and biometric categorisation: a duty to inform Does your chatbot or system use techniques to recognise emotions or to biometrically categorise users? If so, you must explicitly inform your users about this. For most simple webshop chatbots this doesn't apply, but be sure to check if you use a more advanced tool.

What is sufficient in practice?

For the average webshop chatbot, this usually isn't a complicated operation:

• A short, visible notice at the start of the conversation, for example: “You're now chatting with our AI assistant.” Or something more customer-friendly: “Hello! I'm the virtual assistant for [webshop name]. I use AI to answer your questions. Would you rather speak to a member of staff? Just let me know.”
• Avoid the bot actively presenting itself as human (no “I'm Sarah from customer service” without clear AI context).
• Make sure the notice stays visible or is easy to find throughout the conversation, not just a pop-up that disappears immediately.

What if you don't do this?

Fines for breaching the transparency obligation can run up to €15 million, or 3% of global annual turnover. For SMEs and start-ups, the lower of the statutory maximums applies each time, but the risk is real: since 2 August 2026, this is no longer a theoretical obligation, but is now effectively enforceable.

Checklist at a glance

☐ Does your chatbot make it clear that the visitor is talking to AI?

☐ Is that notice visible both at the start and during the conversation?

☐ Does the bot never actively present itself as human?

☐ Do you use emotion recognition or biometrics? If so, has the duty to inform about this been fulfilled?

☐ Are AI-generated texts or images elsewhere on your site clearly marked as such where necessary?

Tick everything off, and you're in good shape for the current obligations.

Not sure whether your chatbot is compliant? Get in touch! In a short conversation, we'll go through together whether your notice meets the requirements and where you might need to make adjustments.

This blog post is intended for informational purposes only and does not constitute legal advice for your specific situation.