TikTok's updated Community Guidelines were released on August 25, 2026 and took effect on September 24, 2026 (TikTok Community Guidelines). If you post AI video, the part that matters is the section called "Edited Media and AI-Generated Content (AIGC)". This guide covers what that section requires, how to turn on the AI-generated label in the app, what happens to realistic AI content that goes up unlabeled, and the short version of the same job on YouTube.
This is not legal advice. Platforms change these pages, so check the linked sources before you rely on them.
What TikTok's guidelines require
The core rule is one sentence: TikTok says it requires "creators to label AI-generated or significantly edited content that shows realistic-looking scenes or people" (TikTok Community Guidelines).
Three definitions in the same section decide whether your video is covered:
- AI-generated content is "any image, video, or audio made or changed by AI," which "can include realistic scenes or artistic styles, like anime, cartoons, or paintings."
- Significantly edited content is media "that makes it seem like someone did or said something they didn't, or alters their appearance so much that they're unrecognizable." TikTok's examples include cutting phrases to change meaning, rearranging or combining clips, and changing speed or adding or removing audio or video.
- Realistic-appearing scenes or people means "content that looks like it could be real, such as AI-generated images that look like real photographs or footage."
So being made with AI is not, on its own, what triggers the requirement. Looking real is.
When you must label
TikTok lists cases where disclosure is needed even when the content "isn't harmful but could be confusing" (TikTok Community Guidelines):
- A face is replaced with someone else's
- AI tools make it look like someone said something they didn't
- A background, object, or person is added or removed in a misleading way
- AI-generated audio mimics the voice of a real person
When you don't need to
The same section says disclosure isn't needed for:
- Small edits like color correction, reframing, or cropping
- Artistic styles, like anime
- Generic text-to-speech narration, "when the TTS isn't a recognizable voice of a known individual"
TikTok's in-app help page puts it the same way: it encourages creators to disclose content that is completely AI-generated or significantly edited with AI, but it requires it "for realistic images, audio, and video" (TikTok, AI-generated content).
What a label can look like
You have two options. TikTok says "you can add your own clear caption, sticker, or watermark," and for AI-generated content "you can also use our AIGC label" (TikTok Community Guidelines). The built-in label is the safer choice because it shows in a consistent place and viewers recognise it.
How to add the AI-generated label on TikTok
TikTok's help page gives these steps (TikTok, AI-generated content).
In the TikTok app:
- Before you post your video, tap More options.
- Turn on AI-generated content.
- Tap Post.
On a computer:
- Click Upload.
- Click Select video and choose your video.
- Click See more.
- Turn on AI-generated content and click Post.
Your video then shows "Creator labeled as AI-generated." Two details from the same page are worth knowing before you post: adding the label "won't affect the distribution of your video," and "you can't remove the label after posting." If you only used TikTok's own effects, you don't need to label, because TikTok shows the effect name on the video.
If you post through the API rather than the app, TikTok's Content Posting API has an is_aigc field in the post info of a Direct Post request, described as "Set to true if the video is AI generated content" (TikTok for Developers; explained in FFmpeg Micro).
Does TikTok label AI video automatically?
Sometimes, but don't count on it. TikTok has a separate "AI-generated" label that it "may automatically apply" to content it identifies as AI-generated, which "may happen when a creator uses TikTok AI effects or uploads AI-generated content that was made on certain platforms" (TikTok, AI-generated content).
The mechanism is C2PA Content Credentials, metadata some AI tools embed in their files. TikTok says C2PA "embeds metadata into content" that lets TikTok and other platforms that use it know when something is AI-generated, and it has tested invisible watermarks on AI content made with TikTok tools and on content uploaded with Content Credentials (TikTok Newsroom, November 2025). In July 2026 TikTok said it had labeled over 3 billion videos as AIGC using Content Credentials, creator labeling tools and invisible watermarking (TikTok Newsroom, July 2026).
The catch is that metadata doesn't always survive. Re-exporting or re-encoding a file in an editor can strip the C2PA data TikTok would read (FFmpeg Micro). If your video needs a label, turn on the toggle yourself rather than hoping detection catches it.
What happens to unlabeled realistic AI content
TikTok says: "Unlabeled content may be removed, restricted, or labeled by our team, depending on the harm it could cause" (TikTok Community Guidelines). It has not published a strike count for missing labels, and the Enforcement section lists repeated rule violations among the grounds for a ban (adtestbench).
Reach is affected too. Under the same section, TikTok lists as ineligible for the For You feed "any realistic-appearing content which isn't yet confirmed to be AIGC or significantly edited content, but presents matters of public importance in a way that could lead to misinterpretation, or cause harm to private figures" (TikTok Community Guidelines). Labeling is what confirms it.
A label is not a free pass. Even labeled, TikTok does not allow AI content that uses a private person's likeness without consent, content made to look like it comes from a real news source, fake crisis events like natural disasters or conflicts, or public figures shown being degraded, endorsing products or taking political stances (TikTok Community Guidelines).
Viewers can now turn AI content down
TikTok added an AI-generated content control to Manage Topics so people can "choose how much AIGC they want to see in their For You feeds." TikTok says it tailors the feed "rather than removing or replacing content in feeds entirely" (TikTok Newsroom, November 2025). The Inquirer gives the path: Settings and privacy, then Content Preferences, then Manage Topics, then AI-Generated Content (Inquirer Technology).
For creators this means some viewers who dial AI down will see less of your work whatever you do. It is not a reason to skip the label. TikTok's help page says the creator label itself doesn't affect distribution (TikTok, AI-generated content), and posting realistic content unlabeled risks removal.
The same job on YouTube
YouTube asks creators to disclose "when they use AI to meaningfully alter or generate photorealistic content." That includes content that makes a real person appear to say or do something they didn't, alters footage of a real event or place, generates a realistic scene that didn't happen, or creates music that is the main focus of the video. Non-realistic content and minor edits don't need it (YouTube Help).
To disclose, go to YouTube Studio and upload as normal. In the Attributes section, under AI use, select Yes if your video meets the requirements. The setting is available in YouTube Studio on a computer or a mobile device. For photorealistic content a label may appear in the video player, and for animated or non-photorealistic content it may appear in the expanded description. YouTube may also add a label itself, for example when a file contains C2PA metadata (YouTube Help).
Disclosure is a separate question from whether a channel can earn money. For that, read YouTube's July 2026 AI monetization rules.
A quick checklist before you post
- Could a viewer think this is real footage? If yes, label it on TikTok and select Yes under AI use on YouTube.
- Does it show or sound like a real person? Label it, and make sure you have consent if it's a private person.
- Is it clearly stylized: animation, illustration, painting? No disclosure is required on either platform, but labeling does no harm on TikTok.
- Is the narration generic text-to-speech? That alone doesn't need a TikTok label. A cloned voice of a known person does.
- Did you re-export the file? Assume any automatic label is gone and use the toggle.
- Does it touch news, a crisis or a public figure? A label won't make misleading content allowed. Rethink it.
Making the video
The label is set where you post, so the platform toggle is the part you control every time. EasyVid renders vertical 9:16 video for TikTok and Shorts as well as landscape and square. You start from a story or script and edit the scenes before anything renders. To see how that works for Shorts and regular uploads, look at the YouTube video maker.