The pressure to produce more content, faster, with fewer resources is real. And AI is the obvious answer: 74% of marketers are already deploying or testing AI-generated content, and 43% plan to increase their investments this year, according to data released today by Validity.
But here’s a paradox for marketers rushing to evolve with AI: When consumers don’t know the content was written by AI, they often prefer it. When they know it, they penalize the brand for it.
A recent Binder A survey of 2,000 British and American consumers highlights the contradiction. Participants saw two articles on the same topic: one written by ChatGPT, the other by a professional writer. Neither was labeled. Of those who had a preference, 56% chose the AI-generated article as more engaging. The AI copy won – when it remained anonymous.
But when participants were told the same content was generated by AI, 52% said they felt less engaged. Same article. Same words. Different reaction.
Validity’s data, drawn from surveys of 500 marketers and 1,000 U.S. consumers, reveals a growing disconnect between how brands use AI and how consumers receive it. On the consumer side, 40% of consumers say they would trust a retailer’s marketing emails less if they knew they were written by AI. Only 25% said knowing an email was created by AI would increase their confidence. And 55% of consumers now make inbox decisions based solely on AI-generated email summaries, without reading the entire message.
Yet only 43% of consumers believe they can actually detect AI-written emails. Others cannot reliably tell the difference, meaning the trust penalty may depend as much on the perception of AI use as on the reality.
What this means for your email strategy
These results highlight a reality that every email marketer needs to consider: AI summaries now function as a primary inbox filter for a growing share of your audience. When 55% of consumers make decisions based solely on summaries, your email copy needs to work on two levels: for the human reading the full message and for the AI rendering a three-line version.
The good news is that writing summaries for AI is not fundamentally different from writing quality emails. Both reward clarity, initial value, and specificity. The bad news is that the attribution model most teams rely on is broken. Fourteen percent of consumers made a purchase based on an AI email summary alone: revenue was entered without them opening the email, meaning no opens or clicks were tracked.
Consider following these steps:
- Review your subject lines and pre-header text through an AI summary lens. Would a reader (or an AI) extract the right value proposition from the first few lines?
- Check your email attribution model. If you only count opens and clicks, you’re likely underestimating the performance of consumers who convert after reading an AI summary.
- Develop a lightweight AI disclosure policy. Bynder data suggests that consumers respond better to honesty than to silence.
A broader trust trend is accelerating
These results are consistent with a longer-term trend. CapgeminiThe global study, which tracks consumer sentiment from 2023 to 2025, found that trust in AI-generated content fell from 73% to 55% in just two years, a decline across all age groups, including Gen Z. And YouGovData from 2026 found that in the markets surveyed, 32% of consumers would trust a brand less if they knew its content was generated by AI, compared to just 15% who would trust it more.
Your customers are searching everywhere. Make sure your brand introduces himself.
The SEO toolkit you know, plus the AI visibility data you need.
Start free trial
Start with
The generational table adds to the complexity. The Bynder survey found that 16- to 24-year-olds were the only age group to find the human-written article more engaging than the AI version – bucking the general trend. While 71% of consumers aged 55 and over want AI-generated content to be disclosed, only 45% of those aged 16 to 24 share the same sentiment. Younger consumers are both more familiar with AI and less bothered by its use – but they are also more likely to penalize brands that get it wrong.
Bynder data also showed what consumers think of brands that use AI: 26% said the brand seems impersonal; 20% said the brand was lazy; and 18% said the brand was not creative. Only 17% said the brand was innovative.
Where is the happy medium?
However, it is possible to navigate without sacrificing the effectiveness of AI. When Bynder asked consumers if they agreed with the statement “I don’t mind if brands use AI to help write copy, as long as the article looks like it was written by a human,” 82% of them agreed, or 41% strongly. And the Validity study found that 35% of consumers said knowing an email was written by AI would not change their trust, suggesting a significant segment prioritizes the outcome over the origin.
For marketers, the way forward is to find a balance between efficiency and trust. The AI content paradox means that with efficiency and scale comes a tax on trust that is activated as soon as a consumer suspects AI involvement. Brands that invest in human review, transparent disclosure, and content that feels truly human rather than automated are best positioned to capture the benefits of AI without paying the trust penalty.
Validity research can be found here And here. (Registration required)
