AI-generated content in advertising: the rules US businesses need to know in 2026

Is it legal to use AI-generated content in ads?

Yes. No US federal law bans AI images, AI voiceovers or AI-written copy in advertising. What is illegal is the same thing that was illegal before AI: misleading people. The tool does not change the standard. The claim does.

The FTC said it directly when it announced its 2024 AI enforcement sweep, Operation AI Comply: there is no AI exemption from the laws on the books. So the useful question is not “can I use AI?” It is “does this AI content make a viewer believe something that is not true?”

This guide covers what applies to a typical US small business running Meta, Google and website ads. It is general information, not legal advice. For anything high stakes, talk to an advertising lawyer.

Which rules apply to AI content in marketing?

Five sets of rules matter for most US businesses. Only one of them is AI-specific at the federal level, and none of them bans AI outright.

What does the FTC say about AI in advertising?

The FTC treats AI content like any other ad content. If a claim would be deceptive coming from a copywriter, it is deceptive coming from a chatbot. Three areas cause most of the trouble.

1. Fake reviews and testimonials

The FTC’s Rule on the Use of Consumer Reviews and Testimonials took effect on October 21, 2024. It bans reviews and testimonials from people who do not exist or who never used the product. The FTC specifically called out AI-generated fake reviews. It also covers buying reviews, hiding a material connection when employees post reviews, and buying fake followers or views.

The rule lets the FTC seek penalties from businesses that knew, or should have known, the conduct was prohibited. Commentators note the FTC could count each fake review as a separate violation. For a small business, ten invented testimonials is not a rounding error.

2. Claims you cannot back up

AI copy tools write confident sentences. “Clinically proven.” “Results in 7 days.” “Trusted by 10,000 homeowners.” If you cannot prove the claim, you cannot run it. The FTC requires a reasonable basis for objective claims before the ad goes live, not after someone complains.

3. Claims about your own AI

If you sell an AI-powered service, describe it honestly. In Operation AI Comply, the FTC went after DoNotPay for marketing an “AI lawyer” it said was never tested against human lawyers. The settlement included a $193,000 payment. Calling a basic tool “AI” to sound advanced, or promising results it cannot deliver, is the same old deception.

Do I have to disclose AI-generated images in Facebook ads?

For regular commercial ads, Meta does not require you to declare AI use. For political, electoral and social issue ads, it does. Meta’s policy requires advertisers to disclose during ad setup when an ad contains a photorealistic image or video, or realistic audio, that was digitally created or altered to show a real person saying or doing something they did not, a realistic-looking person who does not exist, or an event that did not happen. Minor edits like cropping or color correction are exempt.

Separately, Meta adds its own “AI info” label in the About this ad menu for ads created or significantly edited with its generative AI tools. Meta has said it will also detect ads made with third-party AI tools using industry signals such as C2PA. Ads with AI-generated photorealistic people may be labeled more visibly.

Google Ads

Google’s rule is similar and also focused on election ads. Since July 2024, verified election advertisers must tick an “Altered or synthetic content” box in campaign settings when an ad inauthentically depicts real or realistic-looking people or events. For some YouTube and feed formats, Google generates the disclosure. For other formats, you must add a clear one yourself, for example “This image does not depict real events.”

What is New York’s synthetic performer law?

New York became the first state to require disclosure of AI-generated human performers in commercial ads. The law, General Business Law § 396-b, took effect on June 9, 2026. If you produce an ad and know it includes a synthetic performer, meaning an AI-created person who is not a recognizable real individual, you must disclose it conspicuously.

The law does not define “conspicuous” or prescribe wording. Until guidance appears, a short on-screen line like “Features an AI-generated person” for the whole time the performer appears is a sensible default. If you run national campaigns, it is often simpler to apply that standard everywhere than to split creative by state.

Can I use AI voices for calls and voicemails?

This catches more service businesses than ad rules do. In February 2024 the FCC ruled that AI-generated voices count as “artificial” voices under the Telephone Consumer Protection Act. That means AI voice calls and voicemails to consumers need the same prior consent as any robocall. An AI voice agent that answers inbound calls is a different situation, but outbound AI dialing without consent is a fast way to a lawsuit.

Does the EU AI Act apply to US businesses?

It can. Article 2 of the AI Act covers providers and deployers outside the EU when the output of their AI system is used in the Union. A US company running ads to European customers, or using a chatbot on a site that serves EU visitors, is in scope for the transparency rules.

Article 50 applies from August 2, 2026. For a business that uses AI tools, rather than builds them, the duties are short:

• Deepfakes: if you publish AI-generated or manipulated images, audio or video that look real, disclose that they are artificial. For clearly creative or satirical work, a light disclosure that does not ruin the piece is enough.
• Chatbots: people must know they are talking to an AI unless it is obvious. In practice this is the tool provider’s design job, but you choose the setup, so check it.
• Public-interest text: AI text published to inform the public on matters of public interest needs a disclosure, unless a person reviewed it and someone holds editorial responsibility.
• Timing: the disclosure must be clear and appear no later than the first exposure.

The EU’s 2026 Digital Omnibus delayed the high-risk system rules, not these disclosure duties. Only the machine-readable watermarking duty for tool providers got a short extension for systems already on the market.

How to use AI in your ads without legal risk

You do not need a compliance department. You need a short checklist that every AI-assisted ad passes before launch.

1. 1Never let AI invent proof. No AI testimonials, review quotes, star ratings, customer counts or before-and-after photos. Use real customers, with permission.
2. 2Fact-check every claim. Highlight each number, guarantee or result in the AI draft and match it to a source you can show.
3. 3Keep products accurate. AI can create the scene. It cannot change what the product looks like, what it includes or what it costs.
4. 4Disclose AI people. If a realistic AI person appears, add a visible line. It is required in New York and on political ads, and it protects trust everywhere else.
5. 5Flag political and issue ads. Tick the disclosure box on Meta and Google whenever synthetic media touches politics or social issues.
6. 6Check consent before AI calls. Outbound AI voice calls and voicemails need prior consent under the TCPA.
7. 7Tag EU campaigns. If ads or chatbots reach EU users, label realistic synthetic media and make chatbots identify themselves.
8. 8Keep a simple log. Save the prompt, the tool, the date and who approved each AI ad. If a question comes up, you have the paper trail.

What to watch for the rest of 2026

The direction in the US is mixed. Federal policy is pushing to reduce barriers for AI developers, while states keep adding disclosure laws, especially for political deepfakes and synthetic performers. The FTC’s core position on advertisers has not changed: deception is deception. If you set your ads to the strictest common standard, true claims, real reviews and visible labels on realistic AI people, new rules are unlikely to force a rebuild.

This article is general information for marketers and is not legal advice. Laws and platform policies change. Check the official sources linked below, and speak with a qualified attorney before relying on any of it for a specific campaign.

Frequently asked questions