AI-generated content labels on Meta and TikTok ads

Meta and TikTok rules side by side

Meta labels ads itself and only requires advertisers to disclose AI in political and social issue ads. TikTok requires every advertiser to disclose AI-generated or significantly AI-edited content.

Each row is explained below with the official page it comes from. Every rule here was checked on 12 Oct 2026; policies change, so check the current policy page before you launch.

What Meta's AI info label is

AI info is the label Meta uses to show that content was made or edited with AI. It replaced the older "Made with AI" label in July 2024, according to Meta's newsroom post on labelling AI content.

On ads, Meta's help page on AI-generated images in ads says it labels ad images created or significantly edited with Meta's own generative AI creative features. Meta gives background generation and image generation as examples of significant edits.

The same page says minor changes, such as resizing an image or colour correction, don't get AI info. It also says the tools are rolling out gradually, so not every ad made with Meta's AI features will show the label yet.

When Meta labels an ad

Meta applies AI info to ads it identifies as made or significantly edited with AI. For ordinary product ads, the pages checked show no step where the advertiser has to declare AI use.

The help page is written about ad images. It doesn't describe how video ads are handled, which matters for anyone making AI video ads. Its examples are all image features.

Meta's Advertising Standards were checked for a general AI disclosure rule. The only AI-specific advertiser rule found was the political and social issue disclosure covered in the next section.

The newsroom post says Meta adds AI info to content when it detects "industry standard AI image indicators" that other companies put in content made with their tools. That post is about content on Meta's apps in general and doesn't say whether this applies to ads.

Where the AI info label appears on an ad

The label sits on the About this ad screen, opened from the three-dot menu at the top right of the ad. Meta's help page says it sometimes also appears next to the Sponsored label at the top of the ad.

The UK English version of the page is more specific: for ads made with Meta's AI features, the label moves next to Sponsored when the image includes an AI-generated photorealistic human. For AI ad makers this is the case that matters most, because most UGC-style AI ads show a realistic person.

Political and social issue ads on Meta

Advertisers running social issue, election or political ads must disclose photorealistic images, video or realistic audio made or edited with third-party AI tools in three cases, per Meta's political ads policy:

• It shows a real person saying or doing something they didn't say or do.
• It shows a realistic person who doesn't exist, a realistic event that didn't happen, or altered footage of a real event.
• It shows a realistic event that allegedly happened but isn't a true recording of it.

If Meta finds an ad that should have been disclosed and wasn't, it rejects the ad, and repeated failures can lead to penalties on the advertiser. The same page says that from 1 June 2026 Meta also uses automated detection on these ads and adds an AI Info label in About this ad when it finds third-party AI content.

Separately, Meta's misinformation rules say it may reject an ad that is photorealistic, digitally created or altered, and carries a high risk of misleading people on a matter of public importance.

What TikTok requires for ads with AI content

TikTok allows AI-generated and significantly edited media in ads only if you disclose it. Its ad policy on misleading and false content (last updated April 2026) gives two ways: apply the AIGC label, or add your own clear disclaimer, caption, watermark or sticker.

TikTok defines significant AI modification as change beyond minor tweaks. Its examples:

• Images, video or audio that are completely AI-generated.
• The main subject shown doing something they didn't do.
• The main subject saying something they didn't say, using AI voice cloning.

Under that definition, a fully generated AI UGC ad, an AI avatar reading a script, and a cloned voiceover all need disclosure. If TikTok finds undisclosed AI content, the ad is rejected or restricted. TikTok's advice when you are unsure is to label.

How to disclose in TikTok Ads Manager

In Ads Manager, turn on the toggle "This ad contains AI-generated content" while creating the ad. TikTok's guide to ad disclaimers (last updated September 2025) sets two limits:

The second rule catches people who test many versions. If you duplicate a campaign to test a new hook, check the toggle on every new ad.

Spark Ads follow the organic rules

Spark Ads run from organic posts, so TikTok's disclaimer guide says they must follow the Community Guidelines on AI disclosure rather than only the Ads Manager toggle.

The Community Guidelines require creators to label AI-generated or significantly edited content that shows realistic scenes or people. Unlabelled content may be removed, restricted or labelled by TikTok.

In practice, if a creator posts an AI ad that a brand will boost as a Spark Ad, the creator labels the post when it is published. The post carries the label into the ad.

TikTok auto labels and Content Credentials

TikTok may add the "AI-generated" label itself. Its Help Center page on AI content says this can happen when a creator uses TikTok AI effects, or uploads AI content with Content Credentials attached (the C2PA standard).

Once TikTok applies an auto label, you can't remove it from the post. If your AI tool writes Content Credentials into the file, expect the label on organic posts and Spark Ads whether or not you toggled it.

What doesn't count as significant AI editing

Both platforms exclude small technical edits. TikTok lists them; Meta gives a shorter list.

Note the difference on backgrounds. TikTok treats removing or modifying a background as minor; Meta treats generating a background with its own tools as significant. Sources: TikTok ad policy, Meta help page.

Content banned even with a label

A label doesn't make every AI ad acceptable. TikTok's Help Center lists AI content it does not allow at all:

• Fake authoritative sources or fake crisis events.
• Public figures shown being bullied, making an endorsement, or being endorsed.
• The likeness of anyone under 18.
• The likeness of an adult private figure used without their permission.

TikTok's ad policy adds that public figures' likeness, audio or visual, may not be used without permission, especially in misleading contexts. The Community Guidelines also ban paid political advertising, including creators paid to make branded political content.

For AI ad makers, the fake authoritative source rule affects the AI expert ad format: a generated doctor or scientist presented as a real authority is the kind of content listed. A generated person presented as a real passer-by in an AI street interview ad needs the same care.

Where the official pages are unclear or disagree

Several points are not settled by the official pages. This is what each page says.

Where a page is unclear, follow the stricter reading: disclose on TikTok, and assume Meta may label a realistic AI person.

Pre-launch checklist for AI ad makers

Run these checks on every AI ad before it goes live. Each one maps to a rule above.

1. List what is AI. Person, voice, scene, product shots, background. If any of these is generated, treat the ad as AI content on TikTok.
2. Check the edit level. Colour, lighting, denoise and background removal are minor on TikTok. Anything generated is significant.
3. TikTok Ads Manager: turn on "This ad contains AI-generated content", and check it again on every duplicated campaign.
4. TikTok Spark Ads: the creator labels the organic post before the brand boosts it.
5. Or add your own disclaimer. On TikTok, an on-screen caption, watermark or sticker also meets the rule. Keep it readable for the full ad.
6. No fake authorities. Don't present a generated doctor, expert or news anchor as a real authority.
7. No public figures without written permission, and never as an endorser in AI content on TikTok.
8. No minors' likeness in AI content, and no adult private person's likeness without permission.
9. Political or social issue topic? On Meta, disclose AI as the political ads policy requires. On TikTok, paid political ads are not allowed.
10. Expect a label on Meta for realistic AI people, and don't write copy that claims the person is a real customer.
11. Check Content Credentials. If your tool writes C2PA data, TikTok may auto-label the post and you can't remove it.
12. Read the current policy page. These pages were checked on 12 Oct 2026 and change often.

How this fits an AI ad workflow

Disclosure is a step in production, not an afterthought. If you make one script in many formats, as in one winning script, made in every ad format, each version needs its own check, because a podcast version and a street interview version raise different issues.

At Anglr, every video gets a QC and compliance review against ad platform and brand rules before it goes live. Creators and agencies who make AI ads at volume can apply as a creator or apply as an agency. For how AI-made and filmed UGC compare beyond labels, see AI UGC vs real UGC.