AI ad disclosure rules in 2026: Meta, TikTok, YouTube, the FTC and the EU

Meta: political ads must disclose, other ads get auto-labels

Meta's social issue, election and political advertising standard requires advertisers to disclose photorealistic images or video, or realistic audio, created or edited with third-party generative AI tools when the ad shows a real person saying or doing something they did not, shows a realistic-looking person who does not exist or an event that did not happen, or alters footage of a real event. Undisclosed ads are rejected, and repeated failures can lead to penalties.

For ads in general, Meta labels content in two ways. Ads made with Meta's own generative AI tools get an "AI info" label when an edit is significant or when the ad contains an AI-generated photorealistic human; minor edits without a photorealistic human get no label. Since June 1, 2026, Meta also detects ads created or edited with third-party AI tools through industry-standard signals and shows the label under "About this ad" in the three-dot menu. Meta says no advertiser action is required for that automatic label.

What this means for a brand running AI UGC on Facebook and Instagram outside politics: Meta does not ask you to tick a box, but a realistic AI presenter may be labeled automatically. Plan for viewers to see that label and write the ad so it still works, for example as a brand voice rather than a claimed customer review.

TikTok: the AI-generated content toggle and the AIGC label

TikTok's ad disclaimer page (last updated September 2025) covers non-Spark ads whose image, video or audio is either completely AI-generated or significantly modified by AI. For those ads, the advertiser turns on the toggle "This ad contains AI-generated content". After submission the toggle stays locked unless you change the video or image, and duplicating a campaign resets it, so check it on every copy.

Spark Ads are organic posts, so they follow TikTok's community rules for AI-generated content (AIGC) instead. TikTok has required creators to label realistic AI-generated images, audio and video since September 2023, using the AI-generated label, a sticker or the caption. In May 2024 TikTok also began auto-labeling AI content uploaded from some other platforms by reading C2PA Content Credentials.

For AI UGC this is straightforward: an ad with an AI actor is completely AI-generated video, so switch the toggle on. If the same video runs as a Spark Ad from a creator account, the organic post needs the AIGC label.

YouTube and Google Ads

YouTube asks creators to disclose realistic content that is AI-generated or AI-altered: a real person saying or doing something they did not, altered footage of a real event or place, a realistic scene that never happened, or AI music that is the main focus. Disclosure is the "AI use" setting in YouTube Studio. Unrealistic content, beauty or lighting filters, AI-assisted scripts and thumbnails, upscaling and cloning your own voice need no disclosure. Labels appear in the player for photorealistic content and in the expanded description otherwise, and YouTube can apply them automatically from its own tools, C2PA metadata or detection.

Google Ads requires election advertisers to tick "Altered or synthetic content" in campaign settings when an ad contains synthetic or digitally altered content. Google generates the disclosure for some formats; for others the advertiser adds clear wording such as "This video content was synthetically generated". Inconsequential edits such as cropping and color correction are exempt. In July 2026 Google added an optional AI label setting for image and video creatives in all ads, noting that rules in the EU, India and New York require AI disclosures on certain ads and that using the setting does not by itself guarantee compliance.

The FTC: fake testimonials, not AI itself

The FTC's Consumer Reviews and Testimonials Rule (16 CFR Part 465), in effect since October 21, 2024, bans reviews and testimonials that misrepresent that the reviewer exists, that they used the product, or what their experience was. The FTC's announcement names "AI-generated fake reviews" as an example. The FTC's Q&A also says there is "no blanket prohibition" on AI-generated avatars, and that using a celebrity avatar without permission violates the rule.

The Endorsement Guides (16 CFR Part 255) add the classic rule for actors: an ad that shows actors posing as ordinary consumers must disclose that they are actors. Endorsements must reflect honest opinions of people who actually used the product, and the advertiser is liable for what endorsers say.

The line for AI UGC follows from both: an AI presenter delivering a brand script is a dramatization. An AI presenter presented as a real customer who used the product and loved it is a fake testimonial. Write scripts in the first case. Our UGC script templates include problem-solution and demo structures that work as brand voice, and the Ad Remake mode rebuilds a proven structure without copying anyone's testimonial.

New York's synthetic performer law and other state rules

New York General Business Law §396-b requires anyone who creates an ad for commercial purposes to conspicuously disclose that the ad includes a synthetic performer, when they know it does. A synthetic performer is a digitally created, human-looking performer made with generative AI that is not recognizable as a real person. Penalties are $1,000 for a first violation and $5,000 for each later one. Ads for expressive works such as films and games, audio-only ads, and AI used only to translate a human performer are exempt. The Governor's office announced on June 9, 2026 that the law is in effect.

Many other state laws target AI in election ads rather than commercial ads. Minnesota, for example, makes it a crime to spread an election deepfake without consent within set periods before an election. Counts of how many states have such laws vary between trackers, so check the current list for the states where your ad runs.

EU AI Act Article 50

Article 50 of the EU AI Act sets transparency duties. Providers of generative AI systems must mark synthetic audio, image, video and text in a machine-readable, detectable way. Deployers that publish deepfakes must disclose that the content is artificially generated or manipulated, with a lighter duty for evidently artistic, creative, satirical or fictional work. Information must be given at the latest at first exposure. The Act generally applies from August 2, 2026. The 2026 Digital Omnibus (Regulation (EU) 2026/1744) gives providers of generative systems placed on the market before that date until December 2, 2026 to meet the marking duty in Article 50(2).

A disclosure checklist for AI UGC ads

Use this before every launch. Our AI ad disclosure guide covers the reasoning behind it, and our AI disclosure policy covers how FireplaceUGC handles it.

1. 01

   TikTok non-Spark ad with an AI actor: switch on "This ad contains AI-generated content", and check it again after duplicating a campaign.

2. 02

   Meta social issue, election or political ad with realistic AI media: disclose it in the ad flow. For other Meta ads, expect a possible automatic "AI info" label.

3. 03

   Ad shown in New York with a synthetic performer: add a conspicuous on-screen disclosure.

4. 04

   Ad served in the EU from August 2026: label deepfake-style content and keep the provenance metadata your tools add.

5. 05

   Every market: never present an AI presenter as a real customer, never use a real person's likeness without written consent, and substantiate every claim. Put approved claims and banned phrases in the UGC brief so every take inherits them.