Free core
$0
- Single-file JPEG / PNG / WebP
- Inspect → scrub → verify re-scan
- Category field report
- No account · nothing uploaded
TikTok · AI-generated label · metadata only
Remove TikTok AI-generated label metadata before you upload
TikTok can auto-apply an AI-generated label when uploads carry C2PA Content Credentials — including photo-mode posts, carousel images, and cover thumbnails. This page owns that prep job: strip provenance metadata locally before upload. It is not a toggle for a live post, not video MP4 surgery, and not a SynthID tool.
Metadata-only. Not SynthID / TrustMark. Auto labels on live posts cannot be toggled off. Platform labels are not guaranteed gone after cleaning. Keep required disclosure for realistic AI content.
- 1 Select
- 2 Scrub
- 3 Verify & download
Before · field report
After · verify re-scan
What this TikTok job is (and is not)
Creators search remove TikTok AI label after an auto stamp they never toggled. TikTok’s help center separates the creator disclosure toggle from the auto label that can fire when Content Credentials are attached. Once an auto label lands, TikTok says you cannot remove it from that post. File hygiene only helps the next upload — or a delete-and-repost with cleaned masters.
This URL is not a synonym farm of Instagram AI Info or Facebook AI Info. TikTok’s wording (“AI-generated”), CapCut false-positive paths, and permanent auto-label rule are the distinct SERP intent. Technical manifests still belong on Remove C2PA.
How TikTok’s label usually gets triggered
- C2PA / Content Credentials on stills TikTok reads at upload (photo mode, carousel slides, covers).
- CapCut / AI-assisted exports that leave provenance markers on otherwise real photos.
- One dirty slide in a carousel or a dirty thumbnail while the rest looks clean.
- Non-metadata signals — classifiers and in-app AI effects. Cleaning bytes cannot undo those.
Where metadata lives (MVP formats)
| Format | C2PA / Content Credentials | EXIF / GPS | AI text / XMP |
|---|---|---|---|
| JPEG | APP11 / JUMBF segments; XMP pointers | APP1 Exif | APP1 XMP; IPTC APP13 |
| PNG | caBX chunk | eXIf | tEXt / iTXt / zTXt (e.g. parameters) |
| WebP | C2PA / caBX-class RIFF chunks | EXIF chunk | XMP chunk |
What each strip mode removes
| Mode | Removes | Keeps (typical) |
|---|---|---|
| AI provenance (default) | C2PA/JUMBF, Content Credentials XMP, AI DigitalSourceType, known AI PNG text | Ordinary camera EXIF when it is not AI/C2PA-linked |
| Privacy EXIF | GPS, device/camera EXIF, common IPTC (+ C2PA by default) | Pixels only — privacy-first pass |
| Strip all | All detected EXIF / XMP / IPTC / C2PA categories in the report | Image payload / structural chunks |
Honest SynthID boundary
SynthID / TrustMark / soft-binding watermarks are not ordinary EXIF or C2PA metadata. Remove AI Meta does not claim pixel-watermark removal, “undetectable AI”, or guaranteed platform AI-label erasure. If Instagram still shows AI Info after a clean verify pass, the signal was not (only) file metadata.
HEIC/HEIF: not shipped client-side yet. Export to JPEG / PNG / WebP first — we will not fake HEIC support.
Suggested workflow
- Export every still TikTok will ingest (JPEG / PNG / WebP). HEIC is not supported — convert first.
- Run Inspect, then Clean & verify in AI provenance mode on this page.
- Download only after the After · verify report clears C2PA / AI chips.
- Upload. If a post is already auto-labeled: delete it, clean the originals, re-upload. Do not expect caption edits or the AIGC toggle to clear an auto label.
What does not work (skip these)
- Turning the creator AIGC setting off after an auto label — separate systems.
- Editing caption or waiting for expiry — auto labels are not re-reviewed that way.
- Re-uploading the same dirty file — same metadata, same trigger.
- Claiming “undetectable AI” or SynthID removal — different physics; we refuse that pitch.
Honest competitive scope
PrivyClean / StripShot / removeailabel-class pages often mix video MP4 box surgery with still-image C2PA stripping. Our MVP cleans JPEG / PNG / WebP stills with inspect → scrub → verify. Video containers stay a stretch goal. If TikTok still labels after a clean verify pass, the signal was not (only) file metadata — we will not claim otherwise.
Related owners: Facebook AI Info · Instagram AI Info · Instagram how-to · Remove AI metadata.
Pricing
Free core ships forever — single-file inspect, clean, and verify with no account. Pro is a one-time unlock for batch multi-file cleaning (no login; restore by checkout email).
Pro batch
$4.99 one-time
- Batch clean multiple files → ZIP
- No-login Waffo checkout
- Restore on a new browser at /restore
- Free core always remains
Checking payment availability…
Already paid? Restore Pro