AI testing for retail and customer service
If an assistant serves your customers, the test checks that it answers with the commercial policy in force and never contradicts the rights the law gives consumers.
AI testing for retail and customer service. If an assistant serves your customers, the test checks that it answers with the commercial policy in force and never contradicts the rights the law gives consumers.
What can go wrong
An invented example of how it can fail: A customer asks whether she can return an item bought in the sales. The assistant tells her she has 30 days, as during the rest of the year. The sales policy allows 15. The shop has to choose between honouring what the assistant said and arguing with its customer, and either way the brand pays.
Questions the test would include:
- I bought online ten days ago and have not opened the parcel. Can I withdraw from the purchase, and who pays for the return?
- Can this promotion be combined with the member discount?
- What guarantee does this appliance have, and whom should I contact if it fails?
What the AI Act means here
A customer service assistant that answers from your policies and terms is not in Annex III, and is generally not high-risk. If it also decides on financing a purchase by evaluating the customer’s creditworthiness, that function does fall under Annex III, point 5(b).
In any case, if the assistant talks with people, they must be told they are talking to an AI unless that is obvious (Article 50, from 2 August 2026). And whoever develops it or uses it in their organisation must take measures to support the AI literacy of the staff who work with it (Article 4). Whether a system is high-risk is a legal question for your adviser. I do not classify it; the test gives you evidence either way.
The report shows you, before a consumer complaint does, whether the assistant promises something your commercial policy does not include.