TikTok Launches AI Content Labelling to Increase Transparency
In May 2024, TikTok began automatically labelling AI-generated content from other platforms using C2PA Content Credentials, adding to its labels for its own AI effects and creator-applied labels.
.avif)
In May 2024, TikTok became the first video-sharing platform to automatically label AI-generated content uploaded from other platforms, using Content Credentials technology from the Coalition for Content Provenance and Authenticity (C2PA). TikTok already labelled content made with its own AI effects and required creators to label realistic AI-generated content. The new system reads metadata attached to images and videos made with other AI tools and adds an "AI-generated" label automatically.
What TikTok announced
On 9 May 2024, TikTok said it would start automatically flagging AI-generated content created on other platforms. Until then, AI content was labelled in two ways:
- Automatically, when it was made with TikTok's own AI effects.
- By creators, who were already required to label realistic AI-generated content. More than 37 million creators had used TikTok's labelling tool.
The gap was AI content made elsewhere, for example with image generators such as DALL-E, and then uploaded to TikTok. The new system closes part of that gap.
How the labels work
Content Credentials attach metadata to images and videos describing how they were made. When content carrying this metadata is uploaded, TikTok reads it and adds an "AI-generated" label automatically. TikTok also said it would:
- Attach Content Credentials to TikTok content, so the information stays with videos when they're downloaded and other platforms can label them too.
- Join the Adobe-led Content Authenticity Initiative, as the first video-sharing platform to do so.
- Launch media literacy resources to help users understand AI-generated content and misinformation.
What the labels can't do
The system only works when the AI tool used has added Content Credentials. Content from tools that don't use the standard, or where the metadata has been stripped, won't be labelled automatically. That's why TikTok still requires creators to label realistic AI-generated content themselves.
What it means for brands and creators
- Label AI content yourself. If you use generative AI in realistic images or video, disclose it rather than relying on automatic detection.
- Expect labels on AI assets. AI-generated product shots or visuals made with tools that support Content Credentials may be labelled automatically when posted.
- Protect trust. For premium brands especially, being open about AI use protects credibility; audiences react badly to undisclosed AI content.
- Brief creators clearly. Include AI disclosure requirements in creator briefs alongside ad disclosure.
Frequently asked questions
Does TikTok label AI-generated content?
Yes. TikTok labels content made with its own AI effects, requires creators to label realistic AI content, and since May 2024 automatically labels AI content carrying C2PA Content Credentials from other platforms.
What are Content Credentials?
A technical standard from the C2PA that attaches metadata to content showing how it was created or edited, including whether AI was used.
Will TikTok catch all AI-generated content?
No. Automatic labelling only works when the content carries Content Credentials, so creators must still label realistic AI content themselves.
Should brands label AI-generated content on TikTok?
Yes. Labelling realistic AI content is a TikTok requirement and protects audience trust.
For more on TikTok strategy, see our TikTok agency services.
