Do AI-made ads hurt a brand, and can a sense of purpose soften the blow? What an International Journal of Advertising study found
Learning that a campaign came out of an AI tool makes people think a little less of the brand behind it. Four experiments with 972 Americans show that a brand seen to put society ahead of profit can win most of that trust back.
Sands et al. find AI-made ads lower a brand’s credibility, but a brand seen to lead for the greater good wins most of it back, in four US experiments.
Marketing in 3 · 2025 in research · No. 4 · Paper published 4 February 2025
Who made the ad on the screen matters a little. What the brand stands for can matter more. Photo: Brooke Cagle on Unsplash.
The short answer
A little, and purpose helps. In four online experiments with 972 US adults, people who read that a coffee brand's new campaign was made with AI judged the brand less credible and liked it less than people told a creative agency made it. When the brand was also described as putting the wider community ahead of profit, most of the credibility gap closed, and that goodwill carried through to intentions to buy.
Evidence level
★★★One study
★★★Several studies
★★★Meta-analysis
Key takeaways
Learning that a campaign was AI-made cut the brand's credibility from 5.20 to 4.81 out of 7.
A brand known for leading for the greater good closed about 70% of the credibility gap between its AI-made and agency-made campaigns.
The same commitment did nothing for campaigns made by people: their credibility sat at about 5.2 with or without it.
These are modest shifts: whether the work was AI-made explained only 2% to 6% of the differences in how people rated the brand.
What the researchers did
Sean Sands, Vlad Demsar, Carla Ferraro and Sam Wilson of Swinburne University of Technology in Melbourne,
Melissa Wheeler of RMIT University and Colin Campbell of the University of San Diego looked at a worry many marketing
teams now share. AI can produce ad content quickly and cheaply, but people are often uneasy when they find out a machine made it.
The team asked whether a brand's reputation for doing right by society could take the edge off that unease.
The setting was an invented US coffee company with more than 5,000 staff. Participants were given a brief account of the
firm's new campaign, made either with AI or by a creative agency. Some versions also described the company as committed
to "leadership for the greater good": putting decisions that help the community and society before making money.
People then rated the brand. Nobody saw an actual ad; the test was about what people think once they know how an ad was
made.
The first experiment measured the AI penalty on its own. The second crossed AI against the purpose statement. The third
used a lighter version of the purpose statement, and the fourth traced it all the way to purchase plans.
Participants
972 US adults recruited on Prolific, in four separate experiments
Brand
A fictitious coffee company, chosen so no one had prior feelings about it
What changed
Whether AI or an agency produced the campaign, and whether the firm was said to put society before profit
Measured
Brand credibility and attitude towards the brand on seven-point scales; purchase intention in the last study
Checks
Separate pretests confirmed people noticed the changes and found every version equally believable
The AI penalty is real but small
In the first experiment, the AI-made version lowered both how believable and reliable the brand seemed and people's
liking for it. The second experiment found a larger drop of the same kind when no purpose statement was present.
Why would it matter who made an ad people never saw? The authors point to earlier work: people value a human hand
behind what brands say to them, and AI raises worries about privacy and about tools whose workings are hidden. Knowing
a machine was involved is enough to prompt some of that suspicion.
Purpose cushions the penalty
The second experiment carries the main finding. For agency-made campaigns, describing the firm as committed to the
greater good changed almost nothing. For AI-made campaigns it made a clear difference to credibility and a smaller one to
liking, enough to shrink most of the distance between the two kinds of campaign.
One detail matters for practice. In this experiment, the purpose-driven brand also disclosed its AI use as part of that
commitment, while the other AI brand chose not to. The third experiment dropped any mention of disclosure and simply said
the firm wanted its AI to do good for society at large. Credibility and liking still rose.
Liking for the brand, by who made the campaign and whether a purpose was stated (1 to 7 scale)
Liking for the brand, by who made the campaign and whether a purpose was stated (1 to 7 scale)
Group
Value ()
Agency-made, no purpose stated
5.33
Agency-made, purpose stated
5.35
AI-made, no purpose stated
4.51
AI-made, purpose stated
4.87
Experiment 2, 387 US adults, values read from the published figure. The purpose statement mattered only for the AI-made campaign.
From goodwill to buying
The fourth experiment showed every participant an AI-using brand and varied only the purpose statement. The effect on
intentions to buy ran in a chain: the purpose statement raised credibility, credibility raised liking, and liking raised
willingness to buy. There was no shortcut. On this evidence, a purpose claim helps sales only if it first makes the brand
more believable.
AI-made campaign: brand ratings with a lighter purpose statement (1 to 7 scale)
AI-made campaign: brand ratings with a lighter purpose statement (1 to 7 scale)
Group
Value ()
Credibility, no purpose stated
4.28
Credibility, purpose stated
4.65
Liking, no purpose stated
4.56
Liking, purpose stated
5.04
Experiment 3, 193 US adults, every campaign AI-made, no mention of disclosure; values read from the published figure.
The disclosure question
The authors are frank about a tension. Telling people an ad was AI-made can trigger the penalty, so for now some
advertisers may prefer to say nothing unless they can explain what the AI did and why. They also expect regulators to
require disclosure before long, and advise brands to prepare open, plain explanations now. Pairing that openness with a
visible, genuine commitment to the community is, on this evidence, the safer route.
Marketing in 3 steps
Ready. Set. Market.
1
ReadyDecide and prepare
Ask how your audience feels about AI. Before running AI-made creative, ask a sample of customers how they would feel to learn an ad was made with AI. Twenty short interviews or a quick poll of your followers will show whether feelings run strong.
List what your brand does for others. Write down the concrete things your company does beyond profit, such as community programmes, fair sourcing or careful use of customer data. Keep only the ones you can prove.
2
SetBuild it
Link your AI use to that commitment. Wherever you mention AI, add one or two sentences on what it is for and who gains, such as more relevant offers or more staff time for customers. Skip vague claims about innovation.
Draft an honest AI note now. Write a short, plain disclosure for AI-made ads before rules require one. Put it where people will see it, in the caption or on the landing page, not in small print.
3
MarketLaunch and measure
Test the pairing before launch. Show half of a test audience the AI-made ad with your purpose message and half the same ad without it. Compare how believable and likeable each group finds the brand.
Track trust each quarter. Add two questions to your brand tracker: is this brand credible, and does it act for the good of society? Check both after every campaign that uses AI.
Found this useful?
Marketing in 3 is free and ad-free, with new issues every Monday and Thursday. A coffee keeps it that way.
A dent, not a crisis. Revealing AI involvement lowered ratings by between 0.4 and 0.9 on the seven-step scales used.
Purpose pays where AI is involved. For work made by people, the same message added almost nothing.
Belief comes before buying. Any lift in purchase intentions travelled through credibility first, then liking.
A long game. The authors see responsible leadership as something built over time, so expect the cushion to grow with a track record rather than a slogan.
Where it does not work
Descriptions, not ads. Participants read about a campaign; nobody saw or heard real AI-made content.
One invented brand, one category. Established brands with loyal customers, or other product types, may react differently.
Blunt cues. People were told plainly about AI and about the firm's purpose. In daily life many would notice neither.
Purpose and openness overlap. In two of the experiments the purpose statement included disclosing AI use, so the two cannot be fully pulled apart there.
Stated intentions. Willingness to buy was asked, not observed in sales.
Questions marketers ask
Do consumers dislike AI-generated ads?
On average, somewhat. People who knew AI had produced the campaign rated the brand as less credible and less likeable than people told an agency made it, although the gap was modest (Sands et al., 2026).
What is leadership for the greater good?
An approach in which a firm puts the interests of the broader community and society before financial success. Here it was measured by whether people felt the firm makes the world a better place and supports society's long-term welfare.
Should brands tell people when AI made their ads?
The authors see a short-term trade-off: disclosure can provoke the penalty unless it explains AI's role. They expect rules to require it, and a brand that pairs openness with a real commitment to the community fared better in their tests.
Does a purpose message help ads made by people as well?
Not in these experiments. Ratings of agency-made campaigns barely moved when the purpose statement was added; the benefit appeared only where AI was involved.
How does purpose turn into purchases?
Through a chain. The commitment made the brand more believable, that raised how much people liked it, and liking raised their stated willingness to buy.
When a large omnichannel retailer's app went down unexpectedly, shoppers bought less often and spent less, a short-term loss of up to $1.08 million with more erosion estimated later. The surprise was where the money went: store sales fell, because people who browse in the app and buy in person stopped searching, while app buyers simply moved to the website. The app is part of the store.
Across three studies, identity labels given to shoppers pushed them towards matching products more strongly when an algorithm assigned the label than when a person did. Shoppers expected algorithmic predictions to be accurate and identified more with the label as a result. Labelling customers is persuasive, which is also a reason to label carefully.
Source
Sands, S., Demsar, V., Ferraro, C., Wilson, S., Wheeler, M., & Campbell, C. (2026). Easing AI-advertising aversion: how leadership for the greater good buffers negative response to AI-generated ads. International Journal of Advertising, 45(1), 27-47. https://doi.org/10.1080/02650487.2025.2457080
This brief is written for practitioners from the published article. Open access (CC BY-NC-ND 4.0): free to read at the publisher. Numbers are as reported by the authors; read the article for full methods.
Milano et al. tested 98 advertising professionals: generative AI doubled the number of tagline ideas, but only after prompt training, and quality stayed flat.
A detailed "this person does not exist" label cut the sense of being manipulated and raised the sense of artificiality. The two effects cancelled out, except among younger viewers.
Bruno Schivinski is Associate Professor of Marketing at Gdańsk University of Technology. He studies how consumers engage with brands in digital and social media.