TikTok video ineligible for recommendation: AI content rule
TikTok keeps realistic content about public-interest matters out of the For You feed when it is not confirmed as AI or edited. What that means for Sume clips.
One AI-related reason TikTok names for For You feed (FYF) ineligibility is realistic-looking content that is not yet confirmed to be AI-generated or significantly edited, but presents matters of public importance in a way that could lead to misinterpretation, or cause harm to private figures. The page does not describe an appeal process, so the practical step is to make the AI origin clear on the clip.
Read from TikTok's integrity and authenticity guidelines (version 2026H2update) on 2026-09-30. This is a reading of the rule, not a diagnosis of any account.
What exactly is on the FYF-ineligible list?
The "FYF INELIGIBLE" entry in the AI section is a single line about unconfirmed realistic content. Separate "NOT ALLOWED" items cover content that misleads about public importance. The table keeps the three buckets apart.
| Bucket | What the page says | Effect |
|---|---|---|
| Required disclosure | Label realistic AI or significantly edited content | Use the AIGC label or your own caption, watermark or sticker |
| FYF ineligible | Realistic content not yet confirmed as AI or edited, on matters of public importance | Kept out of the For You feed |
| Not allowed | AI or edited content that misleads about a matter of public importance | Prohibited |
Does a clear label avoid this bucket?
The wording turns on content being "not yet confirmed" as AI. Confirming it is the point of disclosure, but the page does not promise a feed outcome for labeled content, and labeled content that misleads is still in the not-allowed bucket.
How do I state AI use on the clip itself?
Pass authored cues (text, start, end) to POST /v1/video-captions to burn overlay copy without speech recognition; see Video captions. The mechanics are in AIGC label or your own caption.
Which Sume generations are most likely to look realistic?
Any of them can. Text-to-video takes a prompt only; image-to-video starts from a first frame you supply with image_url (Video generation). If a clip shows a realistic person or event, treat it as covered by the disclosure rule and keep it away from news-style or crisis framing.
Sources
Related posts
More in Use cases
- Topaz Labs and Adobe: what Sume's image upscale API takes
Adobe completed its Topaz Labs acquisition on Sep 23, 2026, and Topaz stays standalone. If you need an upscale API, here is what Sume's accepts.
- Text to speech reading a confirmation code digit by digit
To make TTS read a code one digit at a time on Sume, write the transcript with the digits separated, pin a Sonic id, and listen before you ship it.
- Twitch clip captions: edit text and timing, burn them in
Twitch is rolling out editable clip captions. For a clip file you edit yourself, Sume burns your own cues with text, start and end, no speech-to-text.
- Twitch vertical clips: crop an older 16:9 clip to 9:16
Twitch Dual Format makes vertical clips going forward. For a 16:9 clip without one, crop a 9:16 slice with video-filter: width 0.316, height 1.
Written by Sume