Correction, 2026-10-02: we re-checked every rule on this page against the law texts and the platforms' own pages. Several platform statements were wrong or unsupported. TikTok's ad policy asks for a label on completely AI-generated content, not only on realistic content, so our FAQ answer "usually not" for obviously fake ads was wrong for TikTok ads. We said Meta asks advertisers to disclose photorealistic AI content; the Meta pages we could check require advertiser self-disclosure only for ads about social issues, elections or politics, and describe labels Meta applies itself to other ads. YouTube's help page no longer mentions the more prominent labels for sensitive topics that we described. We also stated some legal points more broadly than the texts do: the FTC reviews rule took effect on October 21, 2024, the FTC says it can seek civil penalties against "knowing violators", and whether an AI label changes how a testimonial is read is a question for counsel, not a settled rule. Every rule is now quoted from its source with a link and a date, and we added the EU AI Act's deep fake rule, which applies from August 2, 2026.
Correction, 2026-09-30: the first version said no US law requires AI disclosure in ads. That is true at the federal level only: New York's synthetic performer disclosure law has applied since June 9, 2026, and California's takes effect January 1, 2027. Added below.
Short answer: no US federal law requires a "made with AI" label on every ad. Two US states now require a disclosure when an ad uses an AI-generated "synthetic performer" (New York since June 9, 2026; California from January 1, 2027), and in the EU the AI Act requires disclosing deep fakes from August 2, 2026. On top of that, TikTok, Meta and YouTube each have their own label rules, and US consumer-protection law bans deceptive ads and fake testimonials whatever tool made them. This page separates three things that are often mixed up: what the law requires, what platform policy requires, and what is good practice.
This is general information, not legal advice. Laws and platform policies change, and how they apply depends on your ad, your product and where it runs. Check the current text of each rule (every one is linked in Sources, checked 2026-10-02) and ask a lawyer about your specific campaign.
The US federal position, precisely
There is no federal rule that says "an advertisement must be labeled if AI created the visuals." When we searched the Federal Register on 2026-10-02, the only FTC final rules that mention artificial intelligence were the children's privacy (COPPA), impersonation, and consumer reviews rules; none of them requires an AI label on ads. If you read about a finalized federal AI-label rule for ads, check whether it traces to the FTC or the Federal Register before acting on it.
The law that does apply is older. Section 5 of the FTC Act says "unfair or deceptive acts or practices in or affecting commerce, are hereby declared unlawful." The FTC's Policy Statement on Deception (1983) explains how the agency reads that: the Commission "will find deception if there is a representation, omission or practice that is likely to mislead the consumer acting reasonably in the circumstances, to the consumer's detriment," and the representation must be a "material" one. So the question is not "did you use AI" but what the ad leads people to believe.
Two FTC texts make this concrete for AI ads:
- The Endorsement Guides (16 CFR Part 255) set out how the FTC applies Section 5 to endorsements. They say "Endorsements must reflect the honest opinions, findings, beliefs, or experience of the endorser," and that the endorser "could be or appear to be an individual, group, or institution." A connection between endorser and seller that "might materially affect the weight or credibility of the endorsement" and is "not reasonably expected by the audience" must be "disclosed clearly and conspicuously."
- The Rule on the Use of Consumer Reviews and Testimonials (16 CFR Part 465), published in the Federal Register on August 22, 2024 and effective October 21, 2024. It makes it "an unfair or deceptive act or practice" for a business "to 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; or (3) The reviewer's or testimonialist's experience" with it. The FTC's announcement says the rule "addresses reviews and testimonials that misrepresent that they are by someone who does not exist, such as AI-generated fake reviews," and that it lets the agency "seek civil penalties against knowing violators."
So the verified federal position is: no AI-labeling mandate, but rules against fake endorsements and testimonials that AI content can easily run into.
State law: New York now, California from 2027
- New York, General Business Law section 396-b (S.8420-A / A.8887-B, chapter 617 of 2025, signed December 11, 2025). The act takes effect "on the one hundred eightieth day after it shall have become a law," and the Governor's office announced on June 9, 2026 that the law "is now in effect." Any person "engaged in the business of dealing in any property or service who for any commercial purpose produces or creates an advertisement" for it must "conspicuously disclose in such advertisement that a synthetic performer is in such advertisement, where such person has actual knowledge." A synthetic performer is a digitally created asset, made with generative AI or a software algorithm, "that is intended to create the impression that the asset is engaging in an audiovisual and/or visual performance of a human performer who is not recognizable as any identifiable natural performer." The penalty is "one thousand dollars for a first violation, and five thousand dollars for any subsequent violation." The law does not apply to audio ads, to cases where the use of AI "solely involves the language translation of a human performer," to ads for expressive works such as films and video games when the synthetic performer is used the same way as in the work, or to the media (newspapers, TV, streaming services, billboards and so on) that publish the ads.
- California, SB 1050 (Business and Professions Code section 17610, chapter 246 of 2026, approved by the Governor on September 16, 2026). It makes it unlawful "to create and cause to be published in an advertising medium an advertisement that prominently includes a synthetic performer without a clear and conspicuous disclosure that the advertisement includes a synthetic performer." The wording must be "substantially similar to 'this performance features a synthetic performer' or 'no human performer is depicted.'" California's definition covers "a digital figure, voice, or representation" that "creates the realistic impression of the audio, audiovisual, or visual performance of a human performer who is not recognizable as any identifiable natural person," so unlike New York it also reaches voices. "Prominently" covers a synthetic performer in the foreground demonstrating the product, providing or voicing the narration or commercial message, or illustrating or reacting to it. Ads for expressive works (used the same way as in the work) and AI used "solely" for "language translation of a human performer or other accessibility features" are excluded. The bill sets no special start date; under the California Constitution a regular-session statute takes effect "on January 1 next following a 90-day period from the date of enactment," which makes it January 1, 2027.
For a campaign that runs in several states, a cautious approach is to build to the strictest rule that applies and confirm with counsel. Using a real person's likeness is a separate question of consent, covered in synthetic talent and celebrity likeness in ads.
EU: the AI Act's deep fake rule, from August 2, 2026
If your ads run in the EU, Regulation (EU) 2024/1689 (the AI Act) applies. Article 50(4) says: "Deployers of an AI system that generates or manipulates image, audio or video content constituting a deep fake, shall disclose that the content has been artificially generated or manipulated."
- A "deployer" is "a natural or legal person, public authority, agency or other body using an AI system under its authority except where the AI system is used in the course of a personal non-professional activity."
- A "deep fake" is "AI-generated or manipulated image, audio or video content that resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful."
- Where the content is part of "an evidently artistic, creative, satirical, fictional or analogous work or programme," the duty is limited to disclosing that such content exists "in an appropriate manner that does not hamper the display or enjoyment of the work."
- The information must be given "in a clear and distinguishable manner at the latest at the time of the first interaction or exposure."
- Article 113 sets the general date: "It shall apply from 2 August 2026." Chapter IV, which contains Article 50, is not among the exceptions. The consolidated text of 27 July 2026, which includes the Digital Omnibus on AI (Regulation (EU) 2026/1744), still reads this way.
Not every AI ad is a deep fake under that definition. Whether yours is depends on the content and the context, which is a question for a lawyer who knows the countries you run in.
Platform-by-platform label rules
Platform policy is more concrete than the law, and it is what gets an ad rejected or a post relabeled. This is what each platform's own pages said when we checked them on 2026-10-02.
TikTok.
- Ads: TikTok's ad policy on misleading and false content (last updated April 2026) says "Significantly edited media and AIGC content is allowed if the following requirements are met: Apply the AIGC label, or by adding a clear disclaimer, caption, watermark, or sticker of your own." Its examples of content "significantly modified by AI" include "Content that contains images, video, or audio that are completely AI-generated," with no realism test. It adds: "If we identify AI-generated content that has not been disclosed, your ad will be rejected or restricted."
- Organic posts: the Community Guidelines (effective September 24, 2026) say "You must label content that uses AI or includes significant edits to show realistic-looking people or scenes," with TikTok's AIGC label or "your own clear caption, sticker, or watermark." Disclosure is not needed when "Using artistic styles, like anime." "Unlabeled content may be removed, restricted, or labeled by our team, depending on the harm it could cause."
- Automatic labels: TikTok says "we automatically label AIGC made with TikTok AI effects" and "We also label AIGC from certain other platforms by using Content Credentials," the C2PA metadata standard.
Meta (Facebook and Instagram).
- Ads: Meta applies "AI info" labels to "ads created or significantly edited with our generative AI creative tools." In a June 1, 2026 update it said it "will also begin automatically detecting ads created or edited using third-party AI tools through industry-standard signals. When detected, we'll apply an 'AI info' label included in About this ad."
- Ads about social issues, elections or politics: "Advertisers are required to disclose when they use AI to create or alter ads about social issues, elections, or politics in certain cases."
- Organic posts: "We require people to use this disclosure and label tool when they post organic content with a photorealistic video or realistic-sounding audio that was digitally created or altered, and we may apply penalties if they fail to do so."
- We could not confirm on Meta's own pages whether advertisers outside the political and social-issue categories must self-disclose AI content, so check the options in Ads Manager when you upload.
YouTube and Google Ads.
- YouTube uploads: YouTube's help page says "we require creators to disclose when they use AI to meaningfully alter or generate photorealistic content," using the "AI use" setting in YouTube Studio. Creators must disclose content that "Makes a real person appear to say or do something they didn't do," "Alters footage of a real event or place," or "Generates a realistic scene that didn't actually occur." Creators "don't need to disclose non-realistic content that's made with AI." YouTube "may automatically apply an AI label" to content made with its own AI tools, "Content that contains C2PA metadata," and content its systems detect. Creators who "consistently choose not to disclose this information may be subject to manual application of a label, or penalties from YouTube, including removal of content or suspension from the YouTube Partner Program."
- Google Ads: the Misrepresentation policy says "Manipulating media to deceive, defraud, or mislead others is not allowed." For election ads, "Advertisers must disclose all election ads that contain synthetic or digitally altered content by selecting the checkbox in the 'Altered or synthetic content' section in their campaign settings."
The pattern: for organic posts, TikTok, Meta and YouTube all tie the duty to realistic content. TikTok's ad policy goes further and names completely AI-generated content. UGC-style AI ads are realistic by design, so the safe default is to disclose on every platform.
The hard line you never cross
Everything above is about labels. This is about honesty: never present an AI-generated person as a real customer.
An AI avatar saying "I used this for 30 days and here's my before and after" claims an experience that never happened. The FTC reviews rule is aimed at testimonials that materially misrepresent "That the reviewer or testimonialist exists" or that they "used or otherwise had experience with the product." An AI label tells viewers about the medium; it does not make an invented experience true. Do not count on a label to turn a first-person customer claim into something else.
Where it is safer: an AI avatar as an obvious presenter, spokesperson or character delivering brand claims that are themselves true and substantiated. The Endorsement Guides give two examples that are not endorsements: an "obviously fictional dramatization," and a spokesperson who "does not purport to speak from personal experience." Where the line is crossed: an AI persona posing as a real customer relating a personal experience it never had. If your script has the avatar claim lived experience with the product, either cast a real person who actually had that experience (see AI UGC vs real creators for when to do this) or rewrite the claim so the avatar is presenting, not testifying. For anything in between, ask counsel.
What this means for brand teams
The model that keeps you clean is layered, matching the three levels above:
- Legal layer: never fabricate a customer, a testimonial, a review or a result. Disclose synthetic performers where state law requires it (New York now, California from January 1, 2027), and in the EU check whether your content is a deep fake. This is the layer with penalties.
- Platform layer: use each platform's AI disclosure for realistic creative, and on TikTok ads for any completely AI-generated content. TikTok, Meta and YouTube all say they detect some AI content themselves (Content Credentials, C2PA metadata or other industry signals), so self-disclosing gets ahead of an automatic label.
- Trust layer: even where no rule requires it, decide how you will disclose before launch, so the ad never looks like it was hiding something.
Teams that get this wrong usually conflate the layers: they flip the platform toggle and assume that makes any claim fair game. It does not. The toggle discloses the medium; it does nothing for a false claim.
Practical checklist
Before you ship an AI-generated ad, run these:
- Does anyone in the ad claim personal experience with the product? If yes and they are AI, rewrite to a presenter framing or cast a real person.
- Is there a synthetic performer, and where does the ad run? New York (now) and California (from January 1, 2027) require a disclosure in the ad itself; in the EU, check whether the content is a deep fake under the AI Act.
- Is the creative photorealistic, or fully AI-generated? Turn on the AI disclosure on every platform you run (TikTok, Meta, YouTube); on TikTok ads it is required for completely AI-generated content.
- Are there embedded Content Credentials? If your tools write C2PA metadata, leave it in: TikTok, YouTube and Meta say they use such signals to label AI content.
- Any regulated claim? Health, finance, weight loss, earnings, before/after: apply the substantiation and disclosure your category already requires, plus the AI label, and route to counsel.
- Political or social-issue ad? Meta and Google Ads have separate AI-disclosure rules for these ads. Follow those specifically.
- Are your reviews and testimonials real? No AI-generated reviews, no purchased reviews, no invented ratings. The FTC can seek civil penalties against knowing violators of its reviews rule.
- Keep records. Save prompts, model versions and the human sign-off on claims. If a platform or regulator asks who approved this and what it is based on, you want an answer.
FAQ
Is this page legal advice?
No. It is general information about laws and platform policies as their own pages stated them on 2026-10-02. Rules change, and how they apply depends on the facts. Check the current text of each source below and ask a lawyer about your specific ad and the places it will run.
Is it illegal to use AI to make an ad in 2026?
None of the rules on this page bans making ads with AI. They regulate disclosure (New York since June 9, 2026, California from January 1, 2027, the EU AI Act's deep fake rule from August 2, 2026, and the platforms' own labels) and deception (the FTC Act and the FTC's reviews and testimonials rule). Fabricating a customer testimonial or fake reviews is the clearest problem.
Do I have to label an AI ad even if it looks obviously fake?
It depends on the platform. TikTok's ad policy lists "Content that contains images, video, or audio that are completely AI-generated" among the content to label, with no realism test. For organic posts, TikTok and YouTube do not require a label for clearly non-realistic content such as anime or fantasy scenes. When in doubt, TikTok's own ad policy says "it's best to use caution and label it accordingly."
Can I use an AI avatar to read real customer reviews?
Quoting genuine, verifiable reviews through a presenter is different from inventing a customer, as long as it is clear the presenter is quoting customers and not testifying. The Endorsement Guides add that an endorsement "may not be presented out of context or reworded so as to distort in any way the endorser's opinion or experience with the product." Do not have the avatar claim the experience as its own or present itself as the reviewer, and check the script with counsel.
What happens if I skip the platform AI label?
Per the platforms' own pages: on TikTok, an ad with undisclosed AI content "will be rejected or restricted," and unlabeled organic content "may be removed, restricted, or labeled." YouTube may apply a label itself and, for creators who consistently do not disclose, penalties "including removal of content or suspension from the YouTube Partner Program." Meta "may apply penalties" for organic posts without the required disclosure. The platform may also detect and label the content anyway.
Does the EU AI Act require me to label AI ads?
From August 2, 2026, Article 50(4) requires deployers of an AI system that generates or manipulates image, audio or video content "constituting a deep fake" to disclose that the content "has been artificially generated or manipulated." Not every AI ad is a deep fake: the definition covers content that "resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful." Ask a lawyer in the countries you run in about your case.
Disclosure is a workflow step, not an afterthought: build the AI-label toggle and a claims check into your ship checklist. For the honest-casting decisions behind it, see AI UGC vs real creators and the full AI UGC workflow. New to the format? Start with what UGC is.
Sources
- FTC Act, Section 5 (15 U.S.C. 45(a)(1)): Federal Trade Commission Act, ftc.gov, checked 2026-10-02
- FTC Policy Statement on Deception (October 14, 1983): ftc.gov, checked 2026-10-02
- FTC Endorsement Guides, 16 CFR Part 255: eCFR, checked 2026-10-02
- Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465: eCFR; Federal Register, 89 FR 68034, published August 22, 2024, effective October 21, 2024: federalregister.gov; FTC announcement of August 14, 2024: ftc.gov, all checked 2026-10-02
- FTC final rules that mention artificial intelligence: Federal Register search, checked 2026-10-02
- New York GBL 396-b (S.8420-A / A.8887-B, chapter 617 of 2025): NY Senate, NY Assembly, Governor's announcement of June 9, 2026, checked 2026-10-02
- California SB 1050 (chapter 246 of 2026): bill text, bill history; California Constitution, Article IV, Section 8: leginfo, checked 2026-10-02
- EU AI Act, Regulation (EU) 2024/1689, Articles 3, 50 and 113, consolidated text of 27 July 2026: EUR-Lex; Digital Omnibus on AI, Regulation (EU) 2026/1744: EUR-Lex, checked 2026-10-02
- TikTok advertising policy, "Misleading and false content": TikTok Business Help Center, checked 2026-10-02
- TikTok Community Guidelines, "Integrity and Authenticity" (effective September 24, 2026): tiktok.com, checked 2026-10-02
- TikTok, "Supporting responsible, transparent AI-generated content": TikTok Transparency Center, checked 2026-10-02
- Meta, "Expanding GenAI Transparency for Meta's Ads Products" (updated June 1, 2026): about.fb.com, checked 2026-10-02
- Meta, "How Meta Is Preparing for the 2026 US Midterm Elections" (updated June 1, 2026): about.fb.com, checked 2026-10-02
- YouTube Help, "Disclosing use of GenAI content": support.google.com, checked 2026-10-02
- Google Ads policies, "Misrepresentation": support.google.com; "Political content" (election ads, "Altered or synthetic content"): support.google.com, checked 2026-10-02
Try it without signing up: the 8frame canvas opens on a real board. The canvas is free and unlimited; generation is paid from $19/month.