AI UGC or a real creator? Pay for the person when the person's experience is the point. CreatStrat

A company that sells AI UGC wrote this post, so read it with that in mind. The question worth asking first is what the ad has to prove. Price comes second. Some ads only explain a product. Others rest on a person whose experience is the information the shopper is buying, and for those you pay for the person. Use AI UGC to explain, test hooks and show scenes you can't shoot. Use a real person when the shopper needs to believe someone used it.

What is AI UGC, and what is creator UGC?

AI UGC is an ad format where an AI-generated presenter talks to camera about a product, usually cut with product footage; creator UGC is the same format performed by a person you brief and pay. Both borrow the look of a customer filming on a phone: vertical frame, direct address, an ordinary room behind the speaker.

What separates them is who is speaking. An AI presenter presents the product. It can say what's in the box, how the pump works and who the product suits. It can't say it used anything, because nobody was there to use it.

With creator UGC, you write the brief, choose the creator and pay for the clip plus the right to run it. The creator records the clip, and when they say they used the product, that can be true. The brief follows the same structure as any other ad: see what a production-ready ad brief contains.

CreatStrat doesn't offer creator UGC. It's in this post because you can't choose between two formats while describing only one. Neither format is better in the abstract. Each fits some ad jobs and fails at others.

Which one for which ad? A decision table by the job of the ad

Some things a shopper can check before buying: the ingredients, the size, what comes in the box. Economist Phillip Nelson called these search qualities in a 1970 paper in the Journal of Political Economy. The qualities you only learn by using the product, like taste or fit, he called experience qualities. Michael Darby and Edi Karni (1973) added credence qualities, which are hard to judge even after use. Supplements often sit in that third group.

The job of the ad decides which kind of quality it's selling, and that decides whether you need a person.

Pay for a real person when the shopper needs to believe someone's experience; use AI UGC when the ad explains the product, tests a hook, or shows a scene you can't shoot.

Which source fits which ad job. CreatStrat's reasoning from the search, experience and credence framework, not performance data.
Job of the ad	What the shopper needs to believe	Better source	Why
Prove it works on a body like mine (skin or hair, supplements, taste, fit)	That it worked for someone like them (experience or credence qualities)	A creator who used it (check the FTC Endorsement Guides on disclosing a material connection)	The shopper is buying someone's experience, and only a person has one
Explain what it is, how it works, what's in the box	Facts they could check before buying (search qualities)	AI UGC presenter, or a static or carousel	The information is visible, and anyone can state it
Test many hooks on a message you already know works	Nothing new; you need speed and variation	AI UGC	You can change the opening line many times without rebooking anyone. See why many hook tests matter
Show a scene you can't shoot (seasonal, location, scale)	Visual context	AI video or AI UGC	Nobody needs to have been there
You have footage and no time	Reuse	Video mashup	The footage exists; the job is the edit. See video mashups from footage you already have
Any line where someone says "I used this"	A real experience	Only a real customer or creator, or a disclosed actor	See the FTC section below

This table is CreatStrat's reasoning from the search, experience and credence framework. It isn't performance data. No verified study compares AI UGC and creator UGC on Meta results, so we don't claim one wins.

Most products sit in more than one row. A serum has search qualities (2% niacinamide, a pump bottle) and experience qualities (how it feels when you put it on). So decide per ad. The explainer for that serum can use a presenter. The ad built on someone's morning routine needs someone who has one.

A Gap Scan shows which of your ad jobs have no ad yet.

Free Creative Gap Scan →

Free. 3 researched opportunities, 1 customer-avatar gap, and 1 example concept within 24 hours.

Is an AI presenter saying "I tried this" a fake testimonial?

An AI presenter who explains a product isn't giving a consumer testimonial as 16 CFR 465.1 defines it; an AI person who says "I've used this for 30 days" is presenting an experience that never happened.

According to the FTC's rule on consumer reviews and testimonials (16 CFR Part 465, effective October 21, 2024), a business may not create a testimonial that misrepresents that the testimonialist exists or used the product.

Its full name is the Trade Regulation Rule on the Use of Consumer Reviews and Testimonials. The Commission voted 5-0 to issue it, and it was published in the Federal Register on August 22, 2024 (89 FR 68034). Section 465.2(a) says a business may not "write, create, or sell a consumer review, consumer testimonial, or celebrity testimonial that materially misrepresents, expressly or by implication: (1) That the reviewer or testimonialist exists; (2) That the reviewer or testimonialist used or otherwise had experience with the product, service, or business that is the subject of the review or testimonial". A third item covers misrepresenting what that experience was.

What counts as a testimonial? Section 465.1 defines a consumer testimonial as a message that consumers "are likely to believe reflects the opinions, beliefs, or experiences of a consumer who has purchased, used, or otherwise had experience with a product, service, or business."

The rule text doesn't mention AI. The AI wording comes from two other places. The FTC's August 14, 2024 press release describes reviews and testimonials by someone who does not exist, "such as AI-generated fake reviews". Footnote 35 of the rule's preamble warns that "AI tools make it easier for bad actors to pollute the review ecosystem" by producing large numbers of realistic fake reviews quickly and cheaply.

Here is the plain-language reading. A presenter who explains the product isn't giving a consumer testimonial as 16 CFR 465.1 defines it. The script still needs proof: the claims the presenter makes need substantiation under general FTC law. If the presenter says the serum has 2% niacinamide, the serum has to have it.

The lines below are illustrative, written for this post.

Illustrative script lines: a presenter explaining the product, and the same presenter posing as a customer
Presenter line (explains the product)	Fake-customer line (poses as a customer)
"This serum has 2% niacinamide. Here's how you apply it."	"I've used this every morning for a month."
"Here's what's in the box and how the pump works."	"My routine changed completely."

The right-hand lines never say "I'm a customer," and they don't have to. Phrases like "my routine" or "after two weeks" tell the shopper the speaker used the product, and § 465.2(a) covers misrepresentation "expressly or by implication". A presenter script can turn into a testimonial one phrase at a time, so read each line for what it implies about who is speaking.

Human actors are covered by separate FTC guidance. Under the Endorsement Guides (88 FR 48092, July 26, 2023), 16 CFR 255.2(c) says ads presenting endorsements by what are represented 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."

This is educational, not legal advice. Read the rule text or ask counsel.

When do you actually pay a creator?

You pay for the person when the shopper's question is "did it work for someone like me?" Three signs you're there:

The product is judged by using it. Taste, fit, how a supplement fits into a day, how a cream feels or how a shampoo leaves hair. A presenter can describe these. It can't vouch for them.
The script needs the words "I used this." That line has to be true, so it needs a person who did.
The pitch is "someone like you uses this." If the ad's argument is the likeness between the speaker and your customer avatar, you need a speaker who fits that avatar and actually uses the product.

When you go this way, you brief the creator, the creator records the clip, and you pay for the right to run it. A paid creator's testimonial can be completely real. Payment is a connection to the brand, so check the FTC Endorsement Guides on disclosing a material connection, and put the disclosure in your brief. Brief the creator on what they can truthfully say: what to show, which product facts to state, and that any "I used this" has to describe their own use. Rates and usage rights are covered in what a finished Meta ad costs by source, and the brief itself follows what a production-ready ad brief contains.

If your whole ad is a person's experience, AI UGC is the wrong tool. That's an odd thing for a company that sells AI UGC to put in writing. It's still the answer.

What does research say about shoppers and AI presenters?

The studies below measure attitudes, trust and intentions. None of them measures Meta ad results, and none tests AI UGC against creator UGC as ad formats.

In a 2023 Journal of Advertising study, consumers struggled to tell virtual influencers from human ones but still held more positive attitudes toward human influencer campaigns (Franke, Groeppel-Klein and Müller, 2023). The same study found that when the virtual influencer fit the product category, advertising effectiveness went up. The match between presenter and product mattered.

Sands, Campbell, Plangger and Ferraro, writing in the European Journal of Marketing in 2022, found that AI influencers received lower source trust than human influencers. The authors caution against rapidly replacing human influencers.

A 2023 qualitative study by Lou and colleagues, also in the Journal of Advertising, found that virtual influencers helped brand image and awareness but lacked the persuasive ability to drive purchase intention. The authors tie that to a lack of authenticity, low similarity to followers and weak parasocial relations, the one-sided bond a follower feels with someone they watch.

Arango, Singaraju and Niininen (2023) studied charity ads, not product ads. Across three experiments, knowing a face was AI-generated lowered donation intentions.

Taken together, shoppers have trouble spotting an AI presenter and trust it less once they know. That's the case for splitting the work by job: a presenter to explain, a person to vouch.

Does Meta label AI UGC ads?

According to Meta's help page "How AI-generated images in ads are identified and labeled on Meta" (Meta Help Center, accessed October 2, 2026), ad images created or significantly edited with generative AI carry an "AI info" label in the ad's three-dot menu, and sometimes next to the Sponsored label.

The page, "How AI-generated images in ads are identified and labeled on Meta" (Meta Help Center, accessed October 2, 2026), says the label sits "on the About this ad screen in the three-dot menu", and that "Sometimes an AI info label will also appear next to the Sponsored label at the top of the ad". Significant edits "could include features like background generation and image generation". Minor edits such as resizing or color correction don't get the label. Social issue, election and political ads are a separate case: advertisers there are already required to disclose AI-created or altered image, video or audio.

The page is about images. It says nothing specific about AI video presenters, so this post doesn't say how Meta treats an AI UGC video ad. Check Meta's current ad help pages before you launch one.

Illustrative mock / ad images

Where the AI info label sits on an ad image.

Meta's help page names two places: the About this ad screen in the three-dot menu, and sometimes next to the Sponsored label.

An image ad in the feed Your brand

How a weekly ad mix uses both

In practice the table describes a weekly mix. Most weeks bring an explain job, a hook test and a scene you can't shoot, which AI UGC covers alongside statics, carousels and mashups. Experience jobs go to a person you brief and pay yourself.

CreatStrat is a software subscription that delivers finished ads every week. It researches, plans the tests, writes the brief, and produces the finished statics, carousels, mashups, AI video, and AI UGC in your weekly mix, each with its Meta copy. Your creative strategist is built in. Its AI UGC presenter presents the product and doesn't pose as a customer. You pay one combined monthly price: $299 / $499 / $899 a month for 5 / 10 / 20 ads a week, billed monthly, cancel anytime (see the plan).

Each ad starts from why it should exist, which is the same logic as the table. One customer put it this way:

Most AI ad tools we've tried can make something that looks like an ad. CreatStrat is different because it starts with why the ad should exist. It researches the customer, finds the opportunity, develops the concept and then creates the asset. The output feels much closer to working with an experienced strategist.

Alex D. · CreatStrat customer

CreatStrat doesn't offer creator UGC, and it doesn't buy media or run campaigns. You keep the creator side and your ad account.

Frequently asked questions

What is AI UGC? AI UGC is an ad format where an AI-generated presenter talks to camera about a product, usually cut with product footage. It looks like creator-style content but presents the product rather than posing as a customer.

Is AI UGC effective compared with creator UGC? No verified study compares the two on Meta ad results. Research on virtual influencers measures trust and intentions: shoppers have trouble spotting them (Franke et al., 2023) and give them lower source trust than human influencers (Sands et al., 2022). Choose by the job of the ad.

Is an AI presenter saying "I tried this" a fake testimonial? Yes, if it claims personal use that never happened. 16 CFR 465.2(a) bars a business from creating a testimonial that "materially misrepresents, expressly or by implication" that the testimonialist "used or otherwise had experience with the product". This is educational, not legal advice.

Do I have to disclose AI in Facebook ads? For social issue, election or political ads, yes: Meta's help page "How AI-generated images in ads are identified and labeled on Meta" (Meta Help Center, accessed October 2, 2026) says advertisers must disclose AI-created or altered image, video or audio. Otherwise the page covers image labels only. Check Meta's current ad help pages before launch.

Does Meta label AI UGC ads? Meta's help page "How AI-generated images in ads are identified and labeled on Meta" (Meta Help Center, accessed October 2, 2026) covers images, not video presenters, so this post doesn't say. For ad images, the AI info label shows in the three-dot menu and sometimes next to the Sponsored label.

When should I pay a creator instead of using AI UGC? Pay a creator when the shopper needs someone's experience: taste, fit, how a supplement or skincare product works for them. You brief and pay the creator, and the creator's "I used this" has to be true.

Start with the jobs that have no ad

Which format to use comes after a simpler question: which jobs in your account have no ad at all? Maybe nothing explains how the product works, or your best message hasn't had a new hook in weeks. The Gap Scan starts there. You share a product page URL and an email.