Authentic advertising: human-made vs AI ads and customer trust
Short answer Authentic advertising shows what a customer will really get: your people, your work, real customers' words. AI can help draft copy and design, and most Americans accept that. Most reject AI-made people and voices, even when labeled. US rules ban made-up customers in testimonials, and New York now requires a label on AI performers.
Want help applying this to your business? Talk it through on a free 30-minute call.
Book a callWhat is authentic advertising?
An ad is authentic when what it shows is true and a customer could check it: the people are your team or real customers, the photos show your own work, the quotes are words someone actually said, and the price is what you charge. It's not a look. A polished ad can be authentic, and a shaky phone video can still mislead.
One test before any ad goes live: if a customer learned exactly how this ad was made, would they feel tricked? If yes, change the ad.
It matters more now that AI can make a smiling "customer" or a spotless kitchen in seconds, and many people doubt they can spot it. Pew Research Center's 2025 survey of 5,023 US adults found 76% say it's extremely or very important to tell whether pictures, videos and text were made by AI or by people, yet 53% are not too or not at all confident they can.
Do people trust AI-generated ads?
Many don't, and it depends on what the AI made. The 2026 Bentley University-Gallup survey of 3,270 US adults found 49% view businesses' use of AI to create ads negatively, 19% positively and 32% neutral. 79% had seen ads in the past 30 days that looked AI-made.
People draw the line between AI helping and AI replacing a person. Share of US adults who call each use acceptable or not when it is clearly disclosed (Gallup):
| AI use in an ad | Acceptable | Unacceptable |
|---|---|---|
| Brainstorm ideas or early drafts | 75% | 25% |
| Create the final text, images or video | 53% | 46% |
| Create people or voices | 38% | 62% |
So a label doesn't rescue an AI person: 62% reject it even with disclosure.
How do different generations see AI ads?
The youngest adults are the most negative. Gallup's share viewing businesses' AI-made ads negatively, by age:
| Age | Negative |
|---|---|
| 18–29 | 66% |
| 30–44 | 51% |
| 45–59 | 38% |
| 60+ | 46% |
Among 18- to 29-year-olds, 73% say AI-made people or voices in ads are unacceptable. IAB's January 2026 study of 505 US consumers aged 16 to 43 agrees: 45% feel positive about AI-generated ads, while the ad executives surveyed guessed 82%. Gen Z respondents were about twice as likely as Millennials to feel negative (39% vs 20%), and 30% of Gen Z called brands that use AI for ads "inauthentic".
For a local business, no age group is safe to ignore. In every bracket, at least 38% view AI-made ads negatively.
Do AI-made ads perform worse?
Not always, and that's the trap. A 2025 working paper by researchers at NYU Stern and Emory University tested visual ads made by human experts, by humans then edited with AI, and fully by AI. NYU Stern's summary reports that fully AI-made ads raised click rates by up to 19% over the experts' ads. Telling viewers an ad was made with AI cut clicks by 31.5%.
A peer-reviewed study in the Journal of Business Research (Kirk and Givi, 2025) ran seven experiments. When people believed an emotional marketing message was written by AI, they were less likely to recommend the business or stay loyal. The penalty was smaller for factual messages and when AI only edited a person's text. The authors trace it to the message feeling inauthentic.
Put together: an AI-made ad can work until people are told or suspect it. The penalty is biggest where trust matters most: personal stories, thank-you notes and anything presented as a real person.
Where AI is fine, and where you need the real thing
Use AI for the parts nobody expects to be personal. Keep anything that claims to be real, real.
| Part of the ad | Use AI? | Why |
|---|---|---|
| Headline and copy ideas | Yes, then check every fact | 75% accept AI for drafts (Gallup) |
| Lighting and background cleanup | Yes | TikTok calls these minor edits; its label rule covers bigger changes |
| Photos of your work, before and after | No, use real photos | They promise a result; an AI version misleads |
| A presenter talking to camera | Avoid, or label it | 62% reject AI people; New York requires a label |
| A customer testimonial | Never AI | FTC rules require real customers or a clear disclosure |
| Your story, a thank-you message | Write it yourself | Emotional messages lose most when seen as AI-written |
Canada's advertising self-regulator, Ad Standards, says AI-made or AI-altered images must stay truthful about a product's benefits and results, and adds:
"A disclaimer cannot correct a misleading impression." (Ad Standards: Influencer Marketing Disclosure Guidelines)
Can I use an AI avatar as a customer in my ad?
Not without saying so, and even then most viewers object (62% in Gallup's survey). The FTC's Endorsement Guides say:
"Advertisements presenting endorsements by what are represented, expressly or by implication, to be "actual consumers" should utilize actual consumers in both the audio and video, or clearly and conspicuously disclose that the persons in such advertisements are not actual consumers of the advertised product." (16 CFR 255.2(c))
The FTC's rule on reviews and testimonials (16 CFR Part 465) goes further. A business may not create a consumer testimonial that misrepresents that the person exists or used the service. For the penalties and what you may ask real customers, see asking for reviews: the rules.
In Canada, clause 7 of the Canadian Code of Advertising Standards requires testimonials to reflect the genuine, reasonably current opinion of the person giving them, based on real experience. Ad Standards' influencer guidelines add that an AI-generated influencer shouldn't give a testimonial about something it can't experience.
Do I have to disclose AI in ads?
Sometimes, by law or by platform rule:
| Where | Rule |
|---|---|
| New York, since June 9, 2026 | A business that creates an ad for its own goods or services, and knows it includes a "synthetic performer" (a computer- or AI-made person, not recognizable as a real performer) must disclose it conspicuously in the ad. $1,000 for a first violation, $5,000 after. Audio-only ads and AI used only to translate a real performer are exempt. |
| TikTok ads | Significantly AI-edited or AI-generated content needs TikTok's AIGC label or your own clear disclaimer, caption, watermark or sticker. |
| Meta ads | For Meta's AI labels, see Meta Advantage+ for a local business. |
| Canada | The Competition Bureau says there is no "AI exception" to the law. Ad Standards' influencer guidelines say AI-made or heavily AI-altered images showing a product's benefits should be disclosed, for example with #MadeWithAI. |
TikTok enforces its rule at review:
"If we identify AI-generated content that has not been disclosed, your ad will be rejected or restricted." (TikTok Ads Policy: Misleading and false content)
New York's rule sits in General Business Law section 396-b, signed December 11, 2025. A pending bill, A11683, would require the label to stay visible for as long as the AI performer is on screen; in November 2026 it is still in an Assembly committee.
Does a label cost you customers? The evidence is split: IAB found 73% of Gen Z and Millennials say knowing an ad was AI-made would raise or not change their likelihood to buy, while the NYU-Emory tests saw clicks fall 31.5%. So label where a law or platform requires it, and keep AI out of the places a label would hurt: people, results and testimonials.
Example
Illustrative cases with invented numbers.
A roofer near Columbus (US dollars)
A video tool offers AI "customer" avatars for $49 a month. An AI homeowner praising the new roof would be a testimonial from someone who doesn't exist. Instead, the owner asks three recent customers for a 20-second phone video, gets written permission, and uses AI only to write ten headlines to test.
A dental clinic in Brooklyn
The clinic wants an AI presenter to explain implants in a $300 video. In New York that ad needs a conspicuous synthetic performer disclosure, and 62% of US adults reject AI people in ads even when disclosed. A 45-second clip of the dentist answering the same question avoids both problems.
A cleaning company in Toronto (Canadian dollars)
An AI image of a spotless kitchen gets clicks in a C$20-a-day ad, but it shows a result no crew produced. The owner swaps in real after photos from jobs and uses AI only to fix lighting and backgrounds.
FAQ
Is it legal to use AI in ads?
Using AI isn't banned in itself. The rules target what the ad shows and claims: invented customers, fake results, missing labels where a law or platform requires one. Canada's Competition Bureau puts it plainly: there is no "AI exception" to the law.
Can customers tell when an ad is AI-made?
Many doubt they can. 53% of US adults aren't confident they can tell AI from human content (Pew, 2025), while 79% say they recently saw ads that looked AI-made (Gallup, 2026).
Do I need to label AI-edited photos?
On TikTok, only bigger edits. It treats lighting, background and noise fixes as minor; fully AI-made images, or a person shown doing or saying something they didn't, need its label.
Different for you?
- You run Meta ads and want to see what its AI changes: Meta Advantage+ for a local business.
- You want real reviews to quote in ads: asking for reviews: the rules.
- You're planning Facebook and Instagram ads: Facebook ads for a local service business.
- You use AI to write your website: does Google penalize AI content?
- You run a med spa with before-and-after photos: med spa ads rules.
- You'd rather have ads made and run for you: Paid Ads Agent.