Large e-commerce companies have been handling customer conversations together with AI agents for quite some time. Some brands say “this is AI” right away. Others still prefer not to highlight it, because customers may be biased against AI and immediately ask for a human agent.
For Ukrainian businesses, this is still largely a brand choice. In the EU, businesses are already required to inform customers when they are interacting with AI. If a similar rule is introduced in Ukraine, companies will also be legally required to make it clear that the customer is speaking with an AI agent rather than a human.
Walmart
Its assistant is called Sparky. On the store page, it is introduced as “your AI assistant.” It has a name, but it also clearly includes the word AI.
Klarna
The assistant was introduced as AI from the start. In its first month, it handled around two-thirds of customer support chats. Customers could still request a human agent.
IKEA
Its assistant is called Billie. The company describes it as an AI-powered bot. Over two years, it handled almost half of the requests it received.
All three follow a similar model. The bot handles questions about products, delivery, and returns. Human agents take over complaints and more complex cases.
In an Intercom survey from September 2026 involving more than 1,000 people, two of the top expectations were being able to easily reach a human and being told upfront that they were talking to AI.
The customer is less likely to feel misled or assume that support simply does not care. They understand that AI can handle some situations, while in others they can request a human and get one.
In Viber, Telegram, and Instagram, people are more likely to expect that they are talking to a real employee.
Some customers ask for a human as soon as they see the word AI.
In one sales experiment involving phone calls, purchases dropped by more than 79% after such a disclosure. So it is understandable why some teams are hesitant.
A conversation can feel warmer without the word AI.
That is why some brands first give their assistant a name, creating a more human-like experience.
Another customer support study found little difference.
For many people, what matters more is whether their issue was actually resolved and how the conversation felt.
What matters most is that your AI and support team work together smoothly and that everyone understands which processes are handled by whom.
Since August 2, 2026, the EU has had a rule requiring businesses to make it clear when a person is interacting with AI. If a bot is communicating directly with a customer, the disclosure should appear in the first message. Fine print at the bottom of a page is not enough. Using a human name or profile photo in a messenger does not replace this disclosure either.
For businesses selling only in Ukraine, this is not yet a legal requirement. For now, each brand can decide for itself and has time to think about how customer disclosure could be implemented if similar regulation is introduced.
Say It Right Away
“Hi, I’m an AI assistant. I can help with your order, delivery, or return. If you’d like to speak with a person, just let me know and I’ll connect you.”
This can make conversations feel more transparent in messaging apps. It also makes sense if some of your customers are based in the EU.
Give the AI Agent a Name to Make the Experience Feel More Human
Like Sparky or Billie.
The risk is that in a messaging app, the customer may assume they are talking to a real employee. For some brands, that risk is acceptable. For others, it is not.
Don’t Disclose It at All
The conversation can feel more natural, and fewer people immediately ask to “speak to an agent.” Some brands choose this approach because they value a warmer customer experience more than upfront disclosure.
The downside is that if the customer later realizes who they were actually talking to, trust in the brand may drop more sharply than it would have if the disclosure had been made from the beginning.
In every scenario, customers should still have a way to reach a human at any time.
Saying “this is AI” does not replace a real support agent when a customer is making a complaint, returning an expensive product, or dealing with any situation that requires a human decision rather than a simple answer.
