Six in ten Americans say they are turned off when they see the word "AI" in ads or product descriptions, according to a survey by WordPress VIP from June 2026. The survey asked 1,200 people and confirms what many companies have been noticing: putting "AI" on your marketing might actually be hurting your business.

The trust numbers are stark. Eighty-six percent of people surveyed said they don't fully trust AI. Only 14% do. That is a problem for any company putting "AI-powered" or "AI-driven" in their product descriptions or ads. Companies assumed the "AI" label would make people excited. Turns out it makes them skeptical instead.

The survey also looked at how long people will chat with a computer before getting tired. Seventy-four percent of people have felt chatbot fatigue — the feeling that talking to a robot helper is wearing them out. On average, people start feeling tired after about 40 minutes of talking to a chatbot. For companies running customer service with AI helpers, that 40-minute mark is useful to know: it tells them when to hand off to a real person before customers get frustrated.

This is a shift from how companies marketed technology before. Cloud companies did not advertise that their software ran on "TCP/IP," whatever that means to a regular customer. Phone app makers did not highlight "LTE-enabled" to get people excited. AI is different — companies have been splashing the word everywhere: in feature names, menu labels, press releases. Some of that is because AI really is new and impressive. Some of it is because every competitor is trying to look cutting-edge. This survey suggests that the trend might be wearing out faster than companies expect.

Here is the important part: people being turned off by the word "AI" does not mean they hate the technology itself. The real question is whether the capability actually helps them do something. When a chatbot wears people out, it is not necessarily because they hate automation — it is because the automation is either badly designed or kept going too long.

For companies building these products, the practical step is simple. Stop putting "AI" in the places customers see it. Keep the technology working behind the scenes. This is an easy, no-cost test to run. Companies that are already smart about this are the ones that separate what their engineers call their AI plans from what they tell the public about them.

Building real trust is harder. Eighty-six percent of people not trusting AI is not something you fix by changing your marketing words. You build trust by having the technology work reliably, day after day, for a long time. This is how people came to trust online banking, cloud storage, and GPS maps — all of which people were nervous about when they first came out. Building that kind of trust took years, not weeks, and it happened because people used the technology and saw it work. Good marketing helped, but consistent experience was what really mattered.

WordPress VIP is the corporate side of Automattic, and they sell services to help companies manage websites and digital marketing. So read this survey knowing they have a business reason to care about these findings. The survey size — 1,200 people — is a reasonable sample, though not huge, and the questions focused specifically on how people react to "AI" in marketing, which might not tell the whole story about AI in different products or for different ages of people. That said, the basic finding matches what other research has found: companies love talking about AI, but customers are becoming less interested in hearing about it.

The real lesson for companies is this: do not hide AI from your users, but look hard at where you are using the word and ask whether it is actually helping you. In a lot of cases, the honest answer is no. And if it is not helping, getting rid of it costs you nothing.