TikTok ad rejected for a fake play button? The interactive rule
TikTok bans fake play buttons, fake close buttons, dead carousel dots and fake CTA pop-ups inside ad video. How to check an AI-made clip for them before review.

Yes, a drawn-in play button can get an ad rejected. TikTok's Misleading and false content policy (last updated April 2026) says all interactive elements within an ad must function as intended, and it names four things not allowed under 'Non-functioning interactive elements': fake video play buttons that do not start playback, fake close buttons that do not dismiss the ad, non-functional carousel indicators such as dots or arrows when no slides exist, and fake call-to-action buttons or pop-ups meant to make viewers think they must interact to proceed.
AI video makes this easy to trip by accident. A generative model asked for a 'social media ad look' may paint a play triangle, a close X or a button into the frame, and it will look decorative to you and fake to a reviewer.
The four named cases
Use the list as a frame-by-frame review sheet.
| Element | Why it fails | Where AI video adds it by accident |
|---|---|---|
| Fake play button | Does not start playback | A play triangle over a thumbnail-like first frame |
| Fake close button | Does not dismiss the ad | An X in a corner of a mock phone screen |
| Dead carousel dots or arrows | Suggest slides that do not exist | Page dots under a single product shot |
| Fake CTA button or pop-up | Misleads viewers into thinking they must interact | A pop-up card with a button inside the scene |
Check a clip with Sume before upload
You do not need to watch every second twice. Use video inspect to pull stills: with no frames field it returns 8 mid-bin stills, and frames: {"at": [0.5, 1.5, 3]} pulls exactly the instants you name, up to 24 per call. Look at each still for the four shapes above, especially in the first three seconds where a hook is built.
If you find one, regenerate that shot with a prompt that says what to avoid, or cover the offending region. Sume's video filter crop op takes fractions of the frame (x, y, width, height) and the encode costs $0.02 per job, but a crop removes the pixels from the whole clip, so it only works if the element sits at an edge.
Steps
- Pull 8 stills (or the instants of your hook) with video inspect.
- Scan for play, close, dots and pop-up shapes.
- Regenerate or crop the clip if you find any.
- Keep the real call to action in the ad's native CTA button, not in the video.
- Re-inspect the replacement before upload.
Caveat
The policy page is about elements that look interactive but are not. A real on-screen button that is genuinely part of a demonstrated app, shown as part of a screen recording, is a different thing, but a reviewer decides, so keep demos clearly framed as a demo.
Sources
Related posts
More in Use cases
- TikTok AI label source: detection, creator label or TikTok's own tool
TikTok says its AIGC labels now show whether a label came from detection, a creator label, or TikTok's AI tools. What each means for a clip you made elsewhere.
- TikTok comparison ad rules: evidence or a disclaimer, no mocking
TikTok allows comparative claims with evidence or a clear disclaimer, and bars before-and-after effect comparisons and attacks on rivals. A checklist.
- TikTok creator-labeled AI tag: who sets it on an API upload
TikTok's Direct Post API has an is_aigc field for the creator-labeled AI tag. Sume's trim, captions and timeline jobs do not set it; your upload step must.
- TikTok in-feed safe zone files: standard, anchor and Arabic RTL
TikTok publishes four in-feed safe-zone downloads: standard, with anchor, Arabic, Arabic with anchor. Which to use, and how to place a clip's text.
Written by Sume