It can be done, but it is the riskiest use of AI video, because the format works by looking like a real customer’s honest opinion. If a synthetic presenter claims to have used a product, shares a “personal result” or is styled to pass as an ordinary customer, you risk misleading viewers, breaching platform rules and attracting regulator attention. A synthetic presenter that is openly a presenter, delivering claims you can substantiate, is far safer.
Why UGC style is different from other AI video
User-generated content (UGC) style ads borrow the look of a real person speaking to a phone camera. Viewers give such videos more trust because they assume a real person has real experience. That trust is exactly what makes a fake one problematic. Two distinct things can go wrong:
- The presenter is fake but presented as real. A viewer cannot tell that the “customer” does not exist.
- The testimonial is fake. Even with a real actor, a script that invents an experience can be a false endorsement.
Using an AI presenter to explain features, as a clearly branded host, is a different case from inventing a customer story.
What platforms say
Platform rules change often, so read the current policy for each platform before you run an ad. As a reference point, YouTube requires creators to disclose realistic AI-generated or altered content, including content that makes a real person appear to say or do something they did not, or that generates a realistic scene that did not occur. It lists non-realistic content and minor aesthetic edits as not needing disclosure, and warns that creators who consistently fail to disclose may face label changes, removal or suspension from the YouTube Partner Program (YouTube Help). That page governs creator content on YouTube, and paid advertising on any platform has its own ad policies, so check those separately. Other platforms, including Meta and TikTok, have their own labelling and ad rules, which this article does not summarise because they change frequently. Verify them directly.
What consumer law says
In Australia, the Australian Consumer Law prohibits misleading or deceptive conduct in trade or commerce, and the ACCC oversees it. A fabricated testimonial or a “customer result” that never happened can fall within that prohibition. The article on advertising law in Australia gives a plain-language overview. This is general information, not legal advice, and you should seek advice for specific campaigns.
Overseas guidance points the same way. The United States Federal Trade Commission’s endorsement guidance says endorsements must reflect the endorser’s honest opinion and actual experience, that endorsers cannot claim to use a product they have not tried, and that fake reviews or endorsements from people who do not exist or never used the product are deceptive. It also says material connections, such as payment or free products, should be clearly and conspicuously disclosed, and that disclosures in video should be made both visually and audibly where the endorsement is both seen and heard (FTC, Endorsement Guides: what people are asking). This is United States guidance, not Australian law, but it shows the direction regulators are taking and is a sensible standard to meet.
Safer and riskier uses
| Use | Risk | Why |
|---|---|---|
| Branded AI host explaining a product, clearly labelled | Lower | No claim to be a customer |
| Animated or stylised character spokesperson | Lower | Not realistic, unlikely to be mistaken for a person |
| Real actor, scripted, labelled as an ad | Lower to moderate | Needs accurate claims and honest framing |
| Realistic AI “customer” sharing a personal result | High | Fabricated experience, likely misleading |
| Synthetic likeness of a real person without consent | High | Consent, likeness and platform issues |
| AI-generated “reviews” shown as real | Very high | Fake endorsements |
A practical checklist
- Decide what the presenter is. Host, character or fictional customer? Never the last without clear labelling.
- Check every claim. If the presenter says “I lost” or “I saved”, the claim must be true for a real person, or must not be framed as personal experience.
- Disclose AI use in the ad, caption or on-screen label where the presenter looks real, and follow the platform’s labelling tools.
- Disclose paid relationships for any real person who is paid or gifted.
- Get consent and licences for likeness, voice and any model or tool you use. See AI video ethics and disclosure.
- Run a brand safety review before publishing, using the AI video brand safety checklist.
- Keep records of scripts, approvals, tools and sources for claims.
- Monitor comments and reports. If viewers feel deceived, respond honestly and adjust.
Trust costs
Even when an ad is technically allowed, being discovered posing a synthetic person as a real customer can damage a brand. Short-form formats reward authenticity, so the sustainable approach is to use real customer stories with permission, and use synthetic presenters where their artificial nature is not a problem. For format ideas, see short-form video ads for TikTok, Reels and Shorts.
Where to go next
Radiance Impact’s Studio produces AI commercials and short-form and social video with brand safety in mind. See /studio or discuss a campaign via /contact.
Sources
- Disclosing use of altered or synthetic content, YouTube Help
- FTC’s Endorsement Guides: What People Are Asking, United States Federal Trade Commission
