AI label rules for TikTok, Instagram, YouTube and Meta in one table

One table of what is reported about AI labels on four platforms as of 2026-10-02, with a source grade for each row so you know which to verify on the platform.

5 min readSume
All posts

Every row below is a report from a third party, graded by how the page was read. None of it is a platform's own wording, so open each platform's rule page before you rely on a row. The table is a snapshot from 2026-10-02 and will age.

What is reported, platform by platform?

Grades: opened means the page text was read in full on 2026-10-02; snippet means only a search result summary was read.

Reported AI label facts (read 2026-10-02)
PlatformReported factSourceGrade
TikTokLabel for synthetic faces, AI voice cloning, photoreal AI backgrounds and product images; four penalty tiersAudit SocialsOpened, third party
TikTok24 hours to add disclosure or appeal on undisclosed commercial content; AI endorsements prohibitedDarkroomOpened, third party
Instagram"AI creator" label renamed "AI generated profile"; reach limited for unlabeled AI personsSocial Media TodayOpened, news outlet
YouTubeAutomatic branded content labeling from September 2026air.ioOpened, third party
YouTubeOn-video AI labels on Shorts; labels said not to affect reach or monetizationVexubSnippet only
Meta31 Advantage+ enhancements, 17 labeled AILeapBuzzSnippet only

What does Sume add to this?

Sume produces files and a job record. It does not apply any platform's label and it does not decide whether your clip needs one. What it gives you is the evidence per clip: the job id, which tool made it, and the date. The linked disclosure-record post lists the fields worth saving.

How should you use the table?

Pick the platform you publish on, open its own page for the row, and replace the third-party source in your records with the platform's. Anything graded snippet should be verified before it appears in a client document. If a row cannot be verified, drop it.

Sources

Related posts

More in Use cases

All Use cases posts

Written by Sume