Testimonials work because people trust other customers more than they trust brands. That's also why faking them is treated so seriously: a fake review borrows trust it never earned. With AI able to produce a convincing "customer" in minutes, regulators have made the line explicit — and the penalties real.
This article explains the rule in plain words and shows a process for using reviews you can stand behind. It isn't legal advice; if you're in a regulated sector, talk to your counsel.
The FTC's rule, in plain words
In August 2024 the US Federal Trade Commission finalised its Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, in force since October 2024. It lets the FTC seek civil penalties — currently more than $50,000 for each violation — from businesses that, among other things:
- create, buy or sell fake reviews or testimonials: reviews attributed to people who don't exist, who never used the product, or that misrepresent the reviewer's experience. AI-generated fakes are explicitly included;
- buy reviews conditioned on what they say — paying or rewarding only positive (or only negative) reviews;
- post insider reviews without saying so, such as reviews by owners, managers or their families presented as independent customers;
- run a "review site" they control while presenting it as independent;
- suppress reviews with unfounded legal threats or intimidation;
- buy or sell fake social proof — bot followers, views or engagement used to misrepresent influence.
The rule reaches whoever writes, creates or sells the fake, not just the brand that benefits. That's why responsible tools refuse to generate testimonial text at all.
Alongside the rule, the FTC's Endorsement Guides still apply: an endorsement must reflect the honest opinion of an actual user, material connections must be disclosed, and the guides cover virtual influencers too.
What this means for AI presenters
An AI presenter is a spokesperson. Spokespeople are fine: they can introduce your product, explain it and make your offer. What they can't do is claim experiences they never had.
- Fine: "Here's how our serum fits into a morning routine."
- Fine: "Here's what Dana wrote about us: '…'" — quoting a real customer's real words.
- Not fine: "I've used this serum for a month and my skin has never looked better."
- Not fine: A synthetic "customer" in a fake interview saying they switched to you.
Adding "AI-generated" to the caption doesn't fix the last two. The problem isn't that the speaker is synthetic; it's that the experience is invented.
A process for genuine testimonials
1. Start from a real review. Google, Facebook, your own site, an email a customer sent you. Keep the source: a link, a screenshot, the date.
2. Get permission when you need it. Public reviews can often be quoted, but asking the customer is good manners and good protection — especially if you show their name.
3. Quote, don't rewrite. Use the customer's words. Trim for length if you must, but never change the meaning, and never add claims they didn't make.
4. Disclose any incentive. If the customer received something for leaving a review — a discount, a free product, an entry into a draw — the ad should say so.
5. Keep results typical. If a review describes an unusual result, either show what customers generally get or don't use it for a claim about results.
6. Make the frame honest. The presenter reads the review as a quote — "here's what a customer wrote" — with the review on screen. Not in the first person, not as if the presenter were the reviewer.
7. Keep the record. Store the review, its source and your permission next to the video, for as long as the ad might run.
How Facefluence enforces this
We built the rule into the product rather than leaving it to good intentions:
- We never write reviews, and Autopilot never makes testimonials.
- The testimonial template only starts from a real review you paste or link.
- You tick a testimonial declaration every time — that the review is genuine, the words are the customer's, and you may use them.
- The presenter reads the review as a quote, and the source is stored beside the video.
- Accounts that use fake testimonials are removed, suspended or closed, without refund.
That might sound strict. It's meant to be. Fake reviews are the fastest way to lose a customer's trust, a platform's approval and, increasingly, a regulator's patience.
The upside of doing it properly
Real testimonials aren't just safer — they're usually better ads. Real customers use specific, slightly unexpected words ("the croissant was still warm at my desk") that no copywriter would invent. Those details are what make viewers believe the rest of your ad.
So collect reviews deliberately: ask happy customers, make it easy to leave a review, and save the good ones as they arrive. Your best ad scripts may already be sitting in your inbox.