Guide · Compliance

Do you have to disclose AI in ads?

Often, yes. Since June 9, 2026, New York requires ads to disclose AI-generated synthetic performers, and TikTok requires a label on AI-generated ads. The FTC forbids testimonials from people who do not exist and expects ads to say when the "customers" shown are not real customers. Meta labels some AI ads itself.

By Stephen Ventura, Founder, VideoPronto · Updated

Made with VideoPronto

I never filmed this · 0:15 · English What they typed: Make an ad where my digital double looks into the camera and admits I never actually filmed any of this.

The short answer, rule by rule

Whether you must label an AI ad depends on where it runs and what the AI person in it appears to be doing. We read each rule below at its primary source on September 23, 2026. This is a plain-English summary, not legal advice — rules change, and the facts of your ad matter, so talk to an advertising lawyer before you run anything you are unsure about.

RuleIn forceWhat it means for an AI video ad
FTC Endorsement Guides, 16 CFR Part 255Revised July 26, 2023An endorsement must reflect the endorser's honest experience. If people are presented as actual consumers and aren't, say so clearly.
FTC Consumer Reviews and Testimonials Rule, 16 CFR Part 465Since October 21, 2024Businesses may not create or run testimonials that misrepresent that the person exists or used the product. Courts can impose civil penalties for knowing violations.
New York synthetic performer law, GBL § 396-bSince June 9, 2026Ads must conspicuously disclose an AI-generated synthetic performer the advertiser knows about. $1,000 for a first violation, $5,000 after.
TikTok ad policyPage updated April 2026AI-generated or significantly AI-edited ads need TikTok's AI label or your own clear disclaimer.
Meta adsPost updated June 1, 2026Self-disclosure required for social issue, election and political ads. Meta labels other AI ads itself in some cases.
YouTube / Google AdsCurrentCreators must flag realistic altered or synthetic uploads. Google requires disclosure in election ads with synthetic content.

What do the FTC rules actually say about AI people?

The FTC does not ban AI presenters; it bans deceiving people about who is vouching for a product. Three passages matter most for AI UGC ads.

  • An endorser can be someone who only appears to be a person. The Endorsement Guides define the endorser as the party whose experience the message appears to reflect, which "could be or appear to be an individual" (§ 255.0(b)). Endorsements "must reflect the honest opinions, findings, beliefs, or experience of the endorser" (§ 255.1(a)).
  • "Actual consumers" must be actual consumers — or you say they aren't. Ads presenting people as 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" (§ 255.2(c)). The Guides' own example is a hidden-camera cafeteria ad: if actors were used, disclose it.
  • A testimonial from someone who doesn't exist is a violation. Part 465 makes it unlawful for a business to create or run a testimonial that misrepresents "that the reviewer or testimonialist exists" or that they used the product (§ 465.2). When it announced the rule on August 14, 2024, the FTC named AI-generated fake reviews as an example.

The FTC's Q&A on the rule answers the AI question directly. The rule has "no blanket prohibition on the use of AI-generated avatars in marketing." Actors portraying testimonialists are not covered by the rule unless the underlying testimonials are fake or false, though that use could still be deceptive under the FTC Act. And in "an obviously fictional dramatization," the actors are not giving testimonials at all. One line is firm: a celebrity avatar used without permission violates the rule if viewers would think the celebrity really gave the testimonial.

How to disclose matters too. Both FTC texts define "clear and conspicuous" as difficult to miss and easy to understand. In a video, that means on screen and spoken when the claim is both. Under Part 465, the disclosure must also appear in each language the claim is made in, so a Spanish ad needs a Spanish disclosure. A disclosure you have to click or hover to see does not count.

What does New York's synthetic performer law require?

New York calls it a first-in-the-nation law, and it is aimed squarely at AI people in ads. The bill was signed on December 11, 2025 and took effect on June 9, 2026, according to the Governor's announcement. The law covers anyone in business who, for a commercial purpose, produces or creates an ad. If they know the ad includes a synthetic performer, they must "conspicuously disclose" it.

A "synthetic performer" is a digital asset made with generative AI or software to look like a human performer "who is not recognizable as any identifiable natural performer." That describes a typical AI UGC presenter exactly. The law exempts audio-only ads, AI translation of a real performer, and promotion of expressive works such as films or games when it is consistent with the work. If your Meta or TikTok ads can reach people in New York, plan for this law.

What do Meta, TikTok and YouTube require?

  • TikTok: its ad policy lets you run AI-generated or significantly AI-edited content if you "apply the AIGC label" or add "a clear disclaimer, caption, watermark, or sticker of your own." Minor edits such as lighting, color or background changes are exempt. Undisclosed AI ads "will be rejected or restricted." In Ads Manager, the AI-generated content disclaimer puts a text label at the bottom of the in-feed video.
  • Meta (Facebook, Instagram): advertisers must disclose photorealistic AI people and events in ads about social issues, elections or politics. For other ads, Meta adds an "AI info" label itself when its own generative AI tools significantly change an image or video. When the ad shows AI-generated photorealistic people, that label sits next to "Sponsored." Meta also says it will detect ads made with third-party AI tools through industry-standard signals and label them in "About this ad."
  • YouTube and Google Ads: creators must disclose realistic altered or synthetic content when they upload. That includes making a real person appear to say something they didn't, or showing a realistic scene that never happened. Using AI only for a script or thumbnail does not need a label. Google Ads requires a clear disclosure in election ads with synthetic content. We did not find an equivalent blanket label rule for ordinary commercial ads there.

What does this mean for an AI UGC ad?

The riskiest AI ad is a familiar one: a real-looking person who doesn't exist, saying "I tried this and it changed everything." Under Part 465 that is likely a testimonial that misrepresents that the testimonialist exists. Here is how the usual setups compare:

SetupRiskWhat to do
Invented person presented as a real customer ("I used it")HighDon't. Rewrite it as one of the setups below.
A real customer's true story, retold by an AI presenter with the customer's permissionManageableKeep it true and typical. Label it on screen, e.g. "Dramatization. AI presenter. Based on a real customer's experience." Add platform labels.
Spokesperson framing: the presenter speaks for the brand, not as a customerLowerMake no personal-experience claims. Add the TikTok label and New York disclosure where they apply.
Obviously fictional: claymation mascot, Pixar-style storyLowerThe FTC Q&A says obvious dramatization is not a testimonial. Platform AI labels may still apply.
Your own digital doubleLowerSay only what you believe and can back up. Label it where platforms require.
Real customer quotes, verbatim, with permissionLowQuote exactly. Disclose any payment or free product they received.

The sample on this page shows the honest version: a digital double of our founder, made with his consent, telling viewers he never filmed any of it.

What do VideoPronto's Terms say?

You own the ads you make and can use them commercially, and you are responsible for how you use them. Section 4.3 of our Terms puts compliance with "advertising laws, platform policies, disclosure requirements for synthetic media" on the customer. It also limits digital doubles to yourself, or to someone who has given documented, explicit consent. Our acceptable-use rules forbid content that is deceptive, including "undisclosed synthetic impersonations of real people" and "fabricated endorsements." They also forbid content that breaks FTC endorsement rules or the policies of the platform where it runs.

In practice: write your AI ads as brand spokespeople, labeled dramatizations of true stories, obvious fiction, or your own double. Put the disclosure on screen and in the voiceover, in the ad's language. For the creative side, see how to make UGC ads with AI.

Frequently asked questions

Is it legal to use AI people in ads?

Yes, with care. The FTC says its reviews and testimonials rule has no blanket ban on AI-generated avatars. What is illegal is using one to present a testimonial that misrepresents that the person exists or used the product. Disclosure rules, such as New York's and TikTok's, still apply.

Can an AI avatar give a testimonial for my product?

Not as a fake customer. The FTC rule on consumer reviews and testimonials, in effect since October 21, 2024, bars testimonials that misrepresent that the testimonialist exists or used the product. An AI presenter can speak for your brand, or retell a real customer's story with permission, clearly labeled.

Does Meta require me to label AI-generated ads?

Advertiser self-disclosure is required for ads about social issues, elections or politics. For other ads, Meta applies an "AI info" label itself when its own generative AI tools significantly change the ad, and says it is starting to detect third-party AI through industry-standard signals.

Does TikTok require an AI label on ads?

Yes. TikTok's ad policy requires ads that are AI-generated or significantly edited with AI to carry TikTok's AI-generated content label or your own clear disclaimer, caption, watermark or sticker. Minor edits such as lighting or background changes are exempt. Undisclosed AI ads can be rejected or restricted.

What is New York's synthetic performer law?

A New York law, in effect since June 9, 2026, that requires anyone who produces an ad for a commercial purpose to disclose conspicuously when it includes an AI-generated synthetic performer they know about. Penalties are $1,000 for a first violation and $5,000 for each one after.

Do I need to disclose AI if I use a digital double of myself?

On TikTok, yes: AI-generated video needs a label whoever it shows. What your double says must still be true and your honest view. New York's law covers AI performers not recognizable as any real person, so a double of you may fall outside it; ask your lawyer.

Who is responsible for disclosure, VideoPronto or me?

You are. Under section 4.3 of VideoPronto's Terms, customers are responsible for how they use their videos, including advertising law, platform policies and synthetic-media disclosure. The Terms also forbid fabricated endorsements and undisclosed synthetic impersonation of real people.

Video, pronto.

Paste a link or type a line. No subscription — about $3.30 a video.