AINEWS
Consumers Say 'AI' in Brand Messaging Is a Turnoff. They Still Click the Links AI Cites.
WordPress VIP, the enterprise arm of WordPress parent Automattic, ran the survey. It also sells the tools brands use to be cited by AI search.
WordPress VIP, the enterprise arm of WordPress parent Automattic, ran the survey. It also sells the tools brands use to be cited by AI search.
A vendor that sells brands the tools to be cited by AI search has published the clearest data yet that the AI label is a turnoff. WordPress VIP, the enterprise arm of WordPress parent Automattic, commissioned a survey of U.S. consumers and found that 60% of respondents say the word "AI" in a brand's messaging puts them off. In the same survey, 42% said AI-generated answers without clear attribution are trusted less than airline fees, confusing privacy policies, and medical bills. The numbers, first reported by TechCrunch, describe a public that has lost confidence in the AI-shaped parts of the internet, even as enterprises race to plant flags in them.
The full sweep of the survey is more pointed than the headline figure. Eighty-six percent of U.S. consumers said they do not fully trust AI and still want to reach original sources, according to the same WordPress VIP report covered by TechCrunch. Nearly three in four said the internet feels "less human" than it did 10 years ago. Thirty-three percent of consumers said clicking through to see an original source is still their top trust signal, and 80% said information on the web should remain openly accessible rather than controlled by a small number of large organizations. Read together, those responses describe a market that is wary of the technology, tired of the medium, and still hungry for the people and pages behind the answers. That is the opportunity, and the trap, for any brand that has been told to add "AI" to its product page, tagline, or banner ad.
The commercial frame matters and is easy to miss. WordPress VIP is the enterprise division of Automattic, the company that runs WordPress.com, and it sells brands the tooling to make their content legible to AI search crawlers, answer engines, and the new class of agentic shopping assistants. The vendor has a direct financial interest in defining AI-era discovery as a problem worth solving, and its survey lands at the moment its customers are signing contracts to fix it. The numbers are usable, but they are one vendor's reading of the room, not a market research consensus. Independent corroboration — from an academic survey lab or a trust research firm — would sharpen the claim.
The survey is also narrower than the press release suggests. The 60% "turnoff" figure is a self-report on whether the word "AI" in a brand's messaging puts respondents off, not whether they will stop buying, switch providers, or remember the brand negatively a week later. A single survey item can overstate behavioral intent, particularly when the question is asked in isolation. The same caveat applies to the "less trusted than airline fees" framing, which is striking but rhetorical. The directionally sound claim is that consumers prefer human-authored and human-attributed content, and the AI label cuts against that preference in their own telling.
The brands responding to this survey are not, on the whole, planning to retreat. Industry reporting on AI search visibility has shifted from a nice-to-have to a line item in 2026 marketing budgets, with chief marketing officers tracking whether their content gets cited inside ChatGPT, Perplexity, and Google's AI overviews. The strategic bet is that the back-end citation matters more than the front-end label. A brand can quietly optimize for AI retrieval, name a human author on the page, and never put "AI-powered" in a banner ad. The survey data supports that split. Eighty-six percent of consumers want to reach original sources, which means the source pages AI engines cite still need to look and feel human-authored. The win is in being the page the AI cites, not in shouting "AI" in the copy.
There is a usable playbook here, and it does not require abandoning AI. Name the human author on every page an AI engine might cite, with a working byline and a link to that person's work. Disclose AI involvement in plain language, in the same place a reader would look for sourcing notes, not in a 40-line footer policy. Link out to original sources inside any AI-generated summary, so the summary points back to a page the reader can verify. These are old practices. They are the practices that survive the AI summary intact, which is what the survey is asking for when it says consumers want to reach the source.
The unresolved question is whether the distrust follows the label or follows the behavior. If consumers say they are turned off by "AI" in messaging but still click through to AI-cited sources, the brand strategy is to bury the label and invest in citation. If they also distrust the cited sources, the problem is structural, and no byline hygiene will close it. WordPress VIP's data leans toward the first reading, because its commercial pitch is that brands can win the second game (citation) while avoiding the first (the label). The numbers it published are consistent with that pitch. Whether the pitch is right is the next survey worth running, ideally by someone who does not sell the answer.