Six in ten U.S. consumers say the word "AI" in brand messaging puts them off, according to a WordPress VIP survey published on 17 June 2026. The study polled 1,200 U.S. consumers and gives weight to something practitioners have been watching: prominent AI labeling may now be damaging rather than building consumer confidence.

The trust data cuts deeper. Eighty-six percent of respondents said they do not fully trust AI, leaving just 14% who do. That matters for any product team embedding "AI-powered" or "AI-driven" into product features, interfaces, or marketing. The assumption that the label signals capability and innovation does not match what consumers actually feel.

The survey also quantifies chatbot endurance. Seventy-four percent of consumers reported experiencing bot fatigue, with an average threshold of roughly 40 minutes before engagement becomes wearing. For enterprise customer service operations running AI-assisted support at scale, that 40-minute ceiling is operationally useful: it defines when user satisfaction begins to drop without human handoff, and gives product teams a concrete benchmark for routing decisions.

Context matters here. The current AI product wave has been marked by heavy front-of-package branding — "AI" appearing in feature names, labels, and announcements at a density with no parallel in earlier platform shifts. Cloud software vendors did not stamp "TCP/IP-powered" on products; mobile app makers did not lead with "LTE-enabled" as a consumer draw. AI has been marketed differently, partly because the capability is genuinely new, and partly because competitive pressure pushed vendors to signal differentiation quickly. This survey suggests that dynamic may wear out faster than many product teams expected.

None of this means the technology is losing relevance — it means the marketing language around it may be. There is a meaningful gap between a consumer who distrusts "AI" as a brand signal and one who distrusts the capability when it actually helps them accomplish something. The 40-minute bot fatigue window is not evidence that people reject automation outright; it is evidence that poorly calibrated or over-extended automation wears them down.

For engineering and product leadership, the takeaway is fairly straightforward. Removing "AI" from consumer-facing messaging while keeping the capability underneath is a low-cost experiment relative to the signal this data offers. Companies most likely to move quickly on this are already separating their internal AI infrastructure plans from their external messaging strategy — a split that mature platform vendors learned to make in earlier technology cycles.

The trust problem is harder to solve. Eighty-six percent low or no trust is not a number that rewording copy will fix. Closing that gap requires consistent, observable reliability over time — the same pattern that built confidence in online payments, cloud storage, and GPS navigation, all of which faced comparable skepticism in their early mass-market years. Those shifts took years, not quarters, and they were driven mainly by accumulated experience rather than messaging.

WordPress VIP, Automattic's enterprise division, has a direct commercial stake in how brands manage web content and digital experience, so the survey should be read with that context in mind. The sample size — 1,200 U.S. consumers — is modest for broader population estimates, and the framing around "AI in brand messaging" may not capture how attitudes shift across product types, age groups, or interaction styles. That said, the directional findings align with separate industry research tracking a growing gap between vendor enthusiasm for AI labeling and consumer willingness to engage with it.

What this data offers practitioners is not a reason to hide AI from users, but a concrete signal to examine where the label appears and whether it is doing useful work. In many cases, the honest answer will be "very little" — and removing it has no downside.