TikTok has labelled 1.3 billion videos: what that means for your ads

TikTok has labelled a lot of video. The TikTok newsroom post of November 19, 2025, read on 2026-10-05, says it had labelled over 1.3 billion videos to date, and that it uses C2PA Content Credentials, which embed metadata into content, to identify AI-generated material.

That is a TikTok-wide number, not an ad number, and the post does not break it down by organic posts and ads. What it does tell an advertiser is that labelling is routine at scale, so assuming nobody will notice is a poor plan.

What the post lists

What an ad account should do

Treat labelling as the default outcome for a realistic AI ad. Generate with a tool such as the Sume video API, then set the platform's AI-generated content label in the upload flow rather than leaving it to detection.

Sume's docs do not say a clip carries C2PA Content Credentials, so do not tell a client the file is self-identifying. Equally, do not assume detection will miss it.

Build a repeatable step

A big number does not tell you what happens to your account, so avoid quoting it as a risk statistic. Use it for what it supports: labelling is common and expected. Sume's docs describe a metadata field on video, image and music jobs that Sume stores on the job and does not send to the provider. I found nothing in the docs about Sume adding, keeping or removing C2PA credentials, watermarks or platform labels, so treat that as undocumented and check the delivered file.

• Add "AI label set?" as a field in your ad checklist, with yes or no and the reviewer's name.
• Store the Sume job id with each creative so you can trace a flagged ad to its source.
• Test one clip before a launch: upload it as a draft, look at how the label reads, and adjust the plan.
• Re-check TikTok's pages before big flights, since this post is almost a year old by 2026-10-05.