Feeling Misled Is a Question AI Ad Teams Need to Ask After the Sale, Consumer Research Suggests

A consumer survey points to a practical risk for brands using AI in advertising: a satisfying purchase may still leave a customer less willing to trust the advertiser if disclosure arrives later. The reaction makes post-sale sentiment worth tracking alongside clicks, conversions and product ratings.

ContentEngine surveyed 446 U.S. adults about an online advertisement for a product they had not planned to buy. Every respondent was told that the ad persuaded them to purchase and that they were satisfied with the product afterward. The difference came when respondents learned that artificial intelligence had created the ad’s words and images.

In one group, the ad disclosed its AI use before the purchase. In the other, respondents learned about the AI after buying the product and were told that the ad had not disclosed it at the time. The later-discovery group reported lower trust and weaker repeat purchase intent, while more respondents said they felt misled.

The findings offer a useful warning for advertising teams evaluating AI creative. Immediate performance can look healthy while the campaign creates a trust problem that appears later, after the consumer learns more about how the message was produced.

The survey used two separate samples of 223 adults. Each sample read one version of the same hypothetical purchase scenario, then answered eight questions about trust, future buying, regret, feeling misled, recommendations, manipulation, disclosure and human review.

The upfront group saw a scenario in which the advertisement clearly stated that its words and images had been created entirely using AI. The later-discovery group saw a scenario in which the ad persuaded the person to buy, the person enjoyed the product, and the AI involvement was revealed afterward.

The strongest differences appeared in three measures: trust in the advertiser, willingness to buy from it again and feeling misled. The survey reported the following top-two response shares:

The gaps were 11.7 percentage points for trust, 12.6 points for repeat purchase intent and 13 points for feeling misled. After adjustment for the survey’s eight comparisons, those three differences remained statistically significant under the study’s stated model.

The numbers describe the people who completed these surveys. They should be read as directional evidence about reactions to the scenarios rather than as a population estimate for all U.S. adults.

A ContentEngine survey of 446 U.S. adults examined reactions to an online ad that persuaded respondents to make an unplanned purchase they later enjoyed. Respondents who learned about the ad’s AI-generated words and images before buying reported higher trust and repeat purchase intent than those who learned about the AI afterward, while later discovery was associated with stronger feelings of being misled. The findings are directional rather than a population estimate, but they suggest that advertising teams should treat disclosure timing, human review and post-sale sentiment as part of campaign performance.

An upfront disclosure gives the consumer information before the purchase decision. That information may influence whether the person engages with the ad, but it also gives the person a chance to decide how much significance to attach to AI involvement.

A later disclosure arrives after the consumer has acted. The purchase has already happened, and the person has already formed an opinion about the brand’s communication. Learning about AI at that point can change the meaning of the earlier interaction. A message that seemed ordinary at the time may later feel incomplete or carefully managed.

Product satisfaction did not erase that reaction in the survey. Every respondent was told that the purchase ended well, yet the later-discovery group still expressed lower trust and weaker interest in buying again. That result matters because advertising teams often judge a campaign by immediate outcomes such as conversion rate or sales volume.

A customer can like the product and still question the advertiser’s conduct. That combination creates a measurement problem for teams that treat a successful transaction as proof that the advertising experience worked. The sale captures one moment. Trust and repeat purchase intent capture what the customer thinks about the brand after more information becomes available.

The scenario also included an explicit statement that the advertisement had failed to disclose its AI use. That wording likely influenced responses about feeling misled. The study therefore shows an association between later discovery and more negative reactions, while leaving the size of the timing effect open to further testing.

Upfront disclosure produced better results on the clearest trust and loyalty measures, but it did not remove discomfort. Among respondents told about the AI-generated ad before buying, 57% said they felt misled a lot or some. About 62% said they felt manipulated at least somewhat.

Several features of the scenario may explain those reactions. The purchase was unplanned, and the advertisement persuaded the respondent to make it. The ad also used AI for all of its words and images, which presents a stronger case than a campaign in which AI only resized an image or suggested several headline options.

The wording may have made the persuasive experience feel aggressive even when the product performed well. Respondents were not asked why they chose their answers, so the survey cannot separate discomfort caused by AI from discomfort caused by the ad’s influence or the unplanned purchase.

That distinction matters for creative decisions. A disclosure may answer the question of how an ad was produced while leaving other questions untouched: Was the claim accurate? Was the targeting appropriate? Did the creative push too hard? Did the brand retain responsibility for the final message?

Human review received almost the same level of support as AI disclosure. Across both groups, 88.3% said it was very or somewhat important to know that a person had reviewed and approved an AI-generated ad before publication. The support was nearly identical across the samples: 87.9% in the upfront group and 88.8% in the later-discovery group.

Disclosure and review serve different purposes. Disclosure gives consumers information about production. Review assigns responsibility for the claims, images and implications that reach the audience. A small label cannot replace a real approval process.

The practical lesson is not that AI-generated advertising inevitably damages trust; it is that brands should not confuse a successful transaction with a successful communication. Disclosure needs to arrive early enough for consumers to make an informed choice, and it should be paired with substantive human accountability for the final message. Because the survey used hypothetical scenarios and nonprobability samples, teams should validate the pattern with their own audience—but they should not wait for a trust problem to make post-sale measurement part of campaign evaluation.

The survey suggests that AI ad teams should ask a question after the transaction: How might the customer feel if the production process became clear later? That question belongs in campaign planning, not only in crisis response.

A campaign review can examine several points in the customer journey:

The last point is easy to miss. A disclosure can technically exist while remaining too small, vague or distant from the ad to affect the consumer’s understanding. A notice placed on a campaign landing page may have little value if most people see the creative in a short social feed and make a decision there.

The relevant standard is timing plus comprehension. Consumers need enough information, early enough, to incorporate it into their decision. A post-purchase explanation may still provide transparency, but it cannot give the customer the same choice as a pre-purchase disclosure.

Advertising teams can turn the survey’s findings into a basic operating process.

Map the AI involvement. Teams should record whether AI generated copy, images, video, voice, audience segments, product recommendations or variations. The disclosure should match the material role AI played. A campaign with fully generated words and images presents a different communication question from one that used software to remove a background.

Place disclosure near the decision. The notice should appear where the consumer encounters the ad, before the action the ad is designed to prompt. The exact format will vary by channel, but a consumer should not need to complete a purchase before learning about a material production choice.

Define human approval. Review should cover factual claims, visual representations, implied promises, audience suitability and the treatment of sensitive topics. The reviewer should have authority to reject or revise the creative rather than merely record a procedural signoff.

Test the wording with ordinary users. Internal teams often understand technical terms that customers do not. Small research sessions can show whether people understand phrases such as AI-generated, AI-assisted or digitally created. Testing can also reveal whether the notice creates concern about authenticity, accuracy or manipulation.

Track the delayed response. Campaign dashboards should include measures that can move after the initial conversion. Brand trust, repeat purchase intent, customer service contacts, unsubscribe rates, complaints and social discussion can expose problems that a click-through report misses.

Keep an audit trail. A record of the model or tool used, prompts or source material, edits, reviewers and approval dates can help teams investigate a complaint. It also gives the brand a way to show that human oversight involved substantive judgment.

AI campaign success should be evaluated across the customer journey, not only at the moment of conversion.

The survey sits within a broader shift toward clearer disclosure of AI involvement. The Interactive Advertising Bureau’s AI Transparency and Disclosure Framework V2 takes a risk-based and materiality-driven approach to deciding when and how advertising disclosure should occur. That approach is useful for teams handling many different AI applications because it avoids treating every technical feature as equally important to consumers.

European Commission guidance on transparency obligations under Article 50 of the EU AI Act also addresses responsibility in situations where an organization deploys an AI system. The guidance says that employees acting under the authority and control of a legal person, such as advertising or creative staff, are not treated as separate deployers in the same way. Legal obligations depend on the activity, jurisdiction and role of the organization, so campaign teams should obtain qualified legal advice for specific deployments.

The practical lesson is broader than any one rule. Responsibility remains with the organization that puts the advertisement into the market. An agency, platform or software provider may be part of the production chain, but the customer experiences the final ad as a brand communication.

The survey supports a clear operational lesson: earlier disclosure was associated with higher trust and repeat purchase intent than later discovery in these scenarios. It also found broad support for both disclosure and human review.

Several limits keep the finding from supporting a stronger causal claim. The survey used two separate nonprobability samples rather than randomly assigning people to the two scenarios. Differences between the groups may therefore reflect unmeasured characteristics, even though the groups did not differ significantly on the measured demographic variables.

The later-discovery scenario also said that the advertisement had not disclosed its AI use. That statement directly describes nondisclosure and may have increased the number of respondents who felt misled. The scenarios were hypothetical, and respondents did not see a real advertisement, spend actual money or use a real product.

The survey focused on ads whose words and images were created entirely using AI. Its results may not apply in the same way to lighter forms of assistance, such as generating image variations, editing a draft or helping a human copywriter. The study also did not measure race, ethnicity or education, and it did not ask respondents to explain their answers.

Those constraints make the results more useful as a warning signal than as a universal forecast. More research could test the same disclosure timing within a randomized experiment, compare different AI roles, examine actual advertisements and follow people after a real purchase.

AI advertising teams have an incentive to focus on what happens at the point of exposure: impressions, engagement, conversion and revenue. The survey suggests that another checkpoint deserves attention after the transaction. Customers may revise their view of the brand when they learn how an advertisement was made.

A satisfying product can protect the customer relationship, but it may not fully protect the advertiser from a disclosure problem. If the brand wants the customer to buy again, recommend the product or trust future claims, the timing and clarity of AI disclosure deserve the same operational attention as the creative itself.

The most useful review question may be simple: If a customer learns about the AI involvement tomorrow, will the customer feel informed, indifferent or misled? The answer will depend on the ad’s claims, the role of AI, the visibility of the disclosure and the quality of human review. Asking the question before launch gives the brand a chance to address the trust cost before the audience does.