Do I need to label AI content on TikTok if I only used AI to write the caption?

Usually not. TikTok's label covers realistic AI-generated or significantly edited scenes, people and audio. AI-drafted caption text doesn't make the visuals synthetic, but the caption still has to be accurate.

Is the AI label the same as the commercial content disclosure?

No. The AI label says how the media was made. Content disclosure says the post promotes something: "Your brand" for your own product, "Branded content" for paid partnerships. A founder's AI-assisted promo often needs both.

If I labelled my organic post, is the ad version covered?

No. Ads follow TikTok's advertising policies and are reviewed separately. Use the "This ad contains AI-generated content" toggle in Ads Manager or a clear on-ad disclaimer, or the ad can be rejected or restricted.

TikTok AI Label for Brand Posts: 2026 Founder Checklist

Q: Does the TikTok AI label reduce views?

TikTok hasn't published data on it. Viewers can adjust how much AI content they see through a Manage topics control TikTok has been testing, so there may be some effect. Skipping the label is riskier, because unlabeled realistic AI content can be removed or restricted, and TikTok can detect and label it anyway.

If you use AI anywhere in your TikTok workflow, you have probably hit the same toggle and wondered whether it applies to you. This guide splits a typical founder post into its parts and puts each one in a bucket: label it, usually exempt, or banned even with a label. It also covers how organic posts differ from paid ads, and the commercial content toggle that people often mix up with the AI label.

What TikTok's rule actually says

TikTok's Community Guidelines tell creators to label "AI-generated or significantly edited content that shows realistic-looking scenes or people." If you don't, TikTok can remove the post, restrict it, or add a label itself, depending on how much harm it could cause.

Three things follow from that wording:

The test is realism. A cartoon mascot or an obviously stylised graphic doesn't trigger it the way a photoreal kitchen scene or a lifelike face does.
"Significantly edited" counts. Real footage or photos that AI has changed in a meaningful way are in scope, not only content made from scratch.
TikTok may label your post for you. TikTok reads C2PA Content Credentials, the metadata many AI tools attach to their output, and applies its own label. It has also started testing invisible watermarks on content made with its own AI tools and on uploads that carry Content Credentials. Saying nothing doesn't keep a post unlabeled.

The founder post checklist

Here is a typical product post, broken into parts.

Real product photos, lightly edited: usually exempt

Photos of your actual product with cropping, colour correction, brightness or noise reduction don't need the label. Those are ordinary edits and the image still shows what exists. This is the safest base for any brand post because the product the viewer sees is the one they would get.

Your real product placed into a generated scene: label it

This is the most common grey area. Your product photo is real, but the marble counter, the morning light and the hand holding the jar were generated. The scene looks realistic, so turn the label on. You aren't admitting the product is fake. You're saying the setting was made.

A fully generated product shot: label it, and check it's accurate

If AI made the product image itself, label it. The bigger risk is accuracy. A generated shot that shows a colour, size or finish you don't sell misleads buyers, and a label won't fix that. Use generated images for mood, and use real photos for the product.

An AI voiceover: label it when it sounds like a real person

The guideline is written around realistic scenes and people, and a natural-sounding synthetic voice reading your script sits close to that line. A narrator over product shots is low risk, but the label costs nothing and removes the question, so turn it on. A voice cloned from a real person without their consent is a different case, covered below.

A generated persona: label it

A consistent AI-generated face presenting your app is a realistic-looking person, so it needs the label. Also make sure the persona never claims to be a real customer with a real experience. A made-up face giving a first-person testimonial about results misleads people, whether or not it's labelled.

AI-written captions and hashtags: usually exempt

The label covers what people see and hear in the post: scenes, people, audio. Using AI to draft your caption text doesn't make the visuals synthetic. What you still own is whether the caption is true. An invented statistic is a problem whoever wrote it.

Banned even with a label

The label is for disclosure. It doesn't make prohibited content allowed. Don't post these, labelled or not:

A public figure's face or voice used without permission, especially when it implies they endorse your product. TikTok's ad policy names false endorsements as identity misuse.
A private person's likeness used without their permission, for example a generated version of a real customer or an influencer you never worked with.
Content that is "misleading about matters of public importance or harmful to individuals," in the Community Guidelines' words. For product brands this usually means fake health, safety or results claims made to look like evidence.

Why the label is the safe default for brand accounts

The costs aren't even. If you leave the label off a post that needed it, TikTok can remove or restrict it, and repeated problems on a business account add up. If you turn it on for a borderline post, the worst case is a small note on the video.

Does the TikTok AI label reduce views? TikTok hasn't published data either way. One real factor is that TikTok has been testing a control in Manage topics that lets viewers choose to see more or less AI-generated content in their For You feed. Some viewers who turned it down may see fewer labelled posts, and others who turned it up may see more. Hiding AI use doesn't avoid this, because Content Credentials and watermarks can label the post anyway, and then the label looks like something you tried to hide.

To protect reach, work on what you control. Lead with your real product in the first second. Use generated scenes as backdrops, not as the subject. Judge a format over a week of posts, not one video, since new accounts often start with very low views no matter what the label says.

Organic posts and paid ads follow different rules

The Community Guidelines apply to organic posts. Paid ads go through TikTok's advertising policies, which are stricter and enforced at ad review. Many agency write-ups describe a tightening of ad disclosure in July 2026. TikTok's own policy page on misleading and false content, which was marked last updated April 2026 when we checked, says:

Significantly edited media and AI-generated content are allowed in ads if they carry TikTok's AIGC label or your own clear disclaimer, caption, watermark or sticker.
If TikTok finds undisclosed AI-generated content, the ad is rejected or restricted.
In Ads Manager, the "This ad contains AI-generated content" toggle adds a text disclaimer at the bottom of the in-feed ad for as long as it runs.

So labelling an organic post doesn't cover a later ad made from the same creative. If you promote a post or rebuild it as an ad, check the ad-side disclosure again during ad setup.

The commercial content toggle is a separate thing

Founders often switch on one toggle and think they're done. There are two separate disclosures:

AI-generated content tells viewers how the media was made.
Content disclosure (commercial content) tells viewers the post promotes something. Pick "Your brand" when you promote your own app or product, which shows a "Promotional content" label. Pick "Branded content" when a third party pays you, which shows "Paid partnership".

A founder posting an AI-assisted video about their own app usually needs both: the AI label for the generated scene or voice, and "Your brand" because it's a promotion.

How to turn it on

In the TikTok app, after you tap Next on the post screen, open More options and turn on the AI-generated content setting. Add the content disclosure setting on the same screen. Do it before you publish.

If you post through a tool built on TikTok's Content Posting API, the same choices are fields on the post request. is_aigc adds the AI-generated label, brand_organic_toggle marks promotion of your own business, and brand_content_toggle marks a paid partnership:

{
  "post_info": {
    "title": "Our app plans your week in 30 seconds #productivity",
    "privacy_level": "PUBLIC_TO_EVERYONE",
    "disable_comment": false,
    "disable_duet": false,
    "disable_stitch": false,
    "is_aigc": true,
    "brand_organic_toggle": true,
    "brand_content_toggle": false
  },
  "source_info": {
    "source": "PULL_FROM_URL",
    "video_url": "https://example.com/post.mp4"
  }
}

If you're checking a scheduling tool, ask whether it sets is_aigc for you or leaves it to you. Social Studio sends an AI-generated-content disclosure with every post it publishes. It builds scenes around your uploaded product photos so the product on screen is the real one, and nothing goes live until you press Post on the review screen, where you also set the caption, privacy, comments, duet and stitch.

Quick checklist before you post

Does any scene, face or voice look or sound real but wasn't recorded? Turn on the AI label.
Does the post promote your own product? Turn on content disclosure and choose "Your brand".
Does it use a real person's face or voice without permission, or show a product feature you don't have? Don't post it.
Is it going to run as an ad? Check the ad-side AI disclosure again in Ads Manager.
Not sure? Label it.