An AI feature can be technically impressive and still make a customer uncomfortable. The next brand challenge is to explain the benefit while giving people a clear sense of control.
A signal from the United States
An AP report reviewed on 9 October describes an AP-NORC poll in which 64% of US adults thought AI was developing too quickly. The poll covered 2,140 adults, surveyed from 24 to 28 September, with an overall sampling margin of plus or minus 2.9 percentage points.
That finding concerns the United States. It should not be presented as a measurement of Vietnamese customer sentiment. For brands here, it is a reason to ask local customers more precise questions.
Research the moment, not only the technology
Ask where AI feels helpful and where it feels intrusive. A customer may welcome a faster product search but want a person to resolve a dispute. Someone might accept a generated concept image yet expect accurate photography when assessing the actual condition of an item. The context determines the expectation.
Use short interviews around real tasks. Show the interaction, ask what the person believes is happening and observe whether they know how to get help. Avoid leading questions that invite either enthusiasm or fear. The goal is an experience the customer understands.
Design visible control
Explain what an assistant can do at the point where it is used. Distinguish a suggestion from a confirmed transaction. Offer a clear path to human support and keep the customer’s progress available during that handover. These are our design recommendations, not conclusions directly measured by the poll.
Knock Knock’s marketing perspective begins with the promise a brand makes. If the message is personal service, an AI feature should make that service more responsive and easier to navigate. Test comprehension and successful resolution alongside engagement. Trust becomes more tangible when the customer can see what happens next and who remains responsible.
Sources & further reading
Reviewed on 9 October 2026. Research findings are attributed above; business recommendations and future scenarios are Knock Knock’s editorial analysis.
Editorial imagery is AI-generated and illustrates the topic; it does not document the named research projects.
Explore practical AI business solutions



