Survey Reveals How AI Advertising Can Leave Consumers Feeling Misled

A survey of 446 U.S. adults suggests that the timing of an AI disclosure can shape how consumers judge an advertiser, even after a purchase goes well.

The finding is straightforward for brands using generative tools in advertising: consumers appear more comfortable when they learn about the AI before they act. People who discovered the AI involvement only after buying a product reported lower trust, weaker interest in buying again and stronger feelings of being misled.

The survey does not prove that disclosure timing alone caused every difference. The two groups saw slightly different scenarios, and the respondents were drawn from separate online samples. Still, the results offer a useful warning for marketing teams that measure an AI-generated ad only by clicks, conversions or immediate sales.

A disclosure shown before a purchase gives consumers information while they are deciding whether to trust the message. A disclosure shown later gives them an explanation after the decision has already happened. Those moments carry different meanings.

Before the purchase, the consumer can factor AI involvement into the decision. Some people may care about the technology, the role of human creators or the possibility of errors. Others may not change their mind at all. The important point is that the consumer gets a chance to make that judgment.

After the purchase, the same information can feel like a correction to the original sales process. In the survey, 70% of respondents in the later-discovery group said they felt misled a lot or some. That compared with 57% of respondents who knew upfront.

The wording of the scenario likely contributed to that result. The later group was told that the advertisement had not disclosed its AI use. That statement directly raised the issue of nondisclosure and may have encouraged respondents to focus on whether the brand had kept something from them. The finding still reflects a realistic consumer concern: learning about automation through a later notice, news story or brand response can make the original ad seem less transparent in hindsight.

The difference in repeat purchase intent is especially relevant to marketers. About 80% of the upfront group said they would be very or somewhat likely to buy from the advertiser again. That figure dropped to about 67% in the later-discovery group. A campaign can therefore deliver a short-term sale while creating a weaker basis for the next sale.

A survey of 446 U.S. adults suggests that consumers respond more favorably when advertisers disclose meaningful AI involvement before a purchase decision rather than afterward. Respondents who learned about AI upfront reported higher trust and repeat-purchase intent, while those told later were more likely to feel misled—even though both groups imagined being satisfied with the product. The findings are limited by separate samples, hypothetical scenarios and differences in the wording of the scenarios, but they point to a practical standard: disclose significant AI use clearly before consumers act, maintain human accountability and match review to the risks of the ad.

ContentEngine surveyed two groups of 223 U.S. adults. Both groups imagined seeing an online ad for a product they had not planned to buy. In both scenarios, the ad persuaded them to make the purchase, and they were satisfied with the product afterward.

The key detail was when they learned about the ad’s production:

The later-discovery group gave less favorable answers on the survey’s clearest measures.

The gaps were meaningful in practical terms. Trust fell by 11.7 percentage points, repeat purchase intent fell by 12.6 points and feeling misled rose by 13 points among people who learned about the AI afterward. Those three differences remained statistically significant after the analysis adjusted for the survey’s eight comparisons.

A good product experience had already been built into both scenarios. That detail matters. The respondents were satisfied with their purchases, yet the later AI disclosure still coincided with a weaker relationship with the advertiser. Satisfaction can support loyalty, but it may not repair a sense that important information was withheld.

Upfront transparency produced better results on trust and repeat purchase intent, but it did not make every concern disappear. Even among respondents who saw the disclosure before buying:

Those responses need careful interpretation. The scenario described an unplanned purchase driven by a persuasive advertisement. The feeling of manipulation may have come from the ad’s persuasive force, the unexpected purchase, the use of AI or a combination of those elements.

The survey asked how people felt, but it did not ask why. It also described an advertisement created entirely with AI, which is a stronger use case than an ad where AI helped resize an image, generate alternative headlines or produce an initial draft for a human writer.

That limitation keeps the result from supporting a simple rule that adding a label will solve every trust problem. A disclosure gives consumers information, but the content still has to make credible claims, use appropriate targeting and respect the audience’s expectations. Brands may need to examine the whole persuasion strategy alongside the AI label.

The most useful lesson is not that every AI-assisted ad needs a dramatic warning; it is that brands should treat disclosure as part of the customer relationship, not as a production footnote. Consumers need enough information to decide what AI involvement means to them before they click or buy, while brands need a human review process that makes someone accountable for claims, images, targeting and omissions. A timely, specific disclosure cannot guarantee approval, but hiding the information creates a trust risk that a successful first purchase may not repair.

Consumers also wanted evidence that a person remained accountable. Across both groups, 88% said it was very or somewhat important to know that a person had reviewed and approved an AI-generated advertisement before publication.

That figure was nearly identical in the two samples: 87.9% among respondents told upfront and 88.8% among those who learned later. The consistency suggests that human oversight is a separate concern from disclosure timing.

A disclosure answers one question: how was the ad made? Human review answers another: who checked the final claims, images and implications before the ad reached the public?

A meaningful review process should cover more than spelling, layout or technical quality. A person should check whether:

A brand that labels an ad but cannot explain who approved it may satisfy a narrow transparency requirement while leaving consumers with the same basic concern: who is responsible for what they just saw?

The survey is part of a broader shift in how consumers and policymakers view AI-generated marketing. Background research from Gallup reported that 79% of Americans had recently seen ads that appeared to be made using AI. As synthetic images, voices and copy become more common, consumers may have fewer reliable visual clues about how an ad was produced.

A Canva study found that 70% of consumers believe it will eventually be impossible to identify AI-generated advertising without disclosure. That research also reported that 56% expect this situation within two to five years. Those figures come from a different study and should not be combined with the ContentEngine results, but they point to the same practical issue: audiences may increasingly depend on clear labels instead of their own ability to spot automation.

Industry guidance is moving in that direction as well. The Interactive Advertising Bureau’s AI Transparency and Disclosure Framework uses a risk-based, materiality-focused approach to determine when and how AI involvement should be disclosed in consumer-facing advertising and marketing.

The European Commission’s guidance on Article 50 of the EU AI Act also addresses transparency obligations around certain AI-generated or manipulated content. Its discussion emphasizes the role of the legal entity operating the system and the people acting under its authority. For advertisers, that reinforces a basic accountability principle: AI tools do not replace the brand’s responsibility for the published message.

Legal requirements differ by market and by the type of AI use. The survey does not establish a universal compliance standard. It does show why companies may want a clear internal policy even when a particular use of AI does not trigger an obvious legal obligation.

The survey’s findings are useful, but they should be read within its limits.

The two groups were separate samples rather than respondents randomly assigned to different versions of one experiment. The samples did not differ significantly on the measured demographic variables, but they could still have differed in attitudes toward technology, advertising or corporate trust. The results describe the people who completed the surveys and should not be projected to all U.S. adults.

The scenarios also differed in more than timing. The later-discovery version explicitly said the ad had failed to disclose its AI use. That language could have influenced feelings of being misled. The study therefore shows an association between the scenario and the responses, rather than isolating disclosure timing as the only cause.

The purchase was hypothetical. No respondent viewed an actual ad, spent money or used a real product. Stated trust and purchase intent can differ from behavior at checkout or after a real product experience.

The survey also tested a specific case in which all words and images were AI-generated. The findings may not apply in the same way when AI has a smaller supporting role. A label that is appropriate for a fully synthetic commercial may be unnecessary or confusing for a minor production task, depending on the audience and the materiality of the change.

These limits make the results less useful as a precise forecast and more useful as a decision signal. They point toward the questions brand teams should ask before publishing.

Brands can turn that signal into a workable review process without making every ad feel like a technical disclosure document.

The notice should appear where the consumer encounters the ad, before the click, purchase or sign-up decision. A post-purchase explanation cannot give the audience the same choice.

The wording should be specific enough to mean something. If an ad’s text and images were generated entirely by AI, the notice should say so plainly. If AI only assisted with background removal or early copy drafts, the brand can describe that narrower role when the distinction matters to the audience’s understanding.

A label buried in a terms page or hidden behind a secondary menu may technically exist while doing little practical work. Placement, contrast, duration and mobile presentation all matter. Consumers should not have to investigate the ad to discover how it was produced.

The brand should record who reviewed the ad, what they checked and which version received approval. That record can help with internal quality control, regulatory questions and later corrections. It also gives the company a way to explain its process if consumers ask.

A low-stakes product image and a health-related claim deserve different levels of scrutiny. Review should become more demanding when an ad involves health, finance, children, sensitive personal data, political persuasion, realistic synthetic people or claims that could materially affect a purchase decision.

Before launch, a brand can show the ad and its disclosure to a small research group. The useful questions go beyond whether people noticed the label. They include whether people understood the AI’s role, whether the ad still felt credible and whether the disclosure changed their sense of control over the decision.

The survey gives advertisers a reason to treat AI disclosure as part of customer relationship management rather than as a production detail. The immediate campaign metric may look healthy while trust and repeat purchase intent weaken later.

That risk becomes more visible when consumers discover the AI use through an outside source. A social post, journalist, employee comment or product review can make a previously invisible production choice feel like withheld information. The resulting reaction may attach to the brand, even when the product itself performs as promised.

A visible, timely disclosure cannot guarantee approval. Some consumers will still dislike AI-generated creative or question whether automation belongs in advertising. It does give them a fair chance to decide how much that concern matters before they act.

For brand teams, the clearest rule from the survey is practical: disclose meaningful AI use before the consumer makes a decision, explain the role in plain language and assign a person responsibility for the final ad. A successful purchase is valuable. Earning confidence for the next purchase requires transparency at the first one.