How to remove Content Credentials in three steps
- Drop one or more images into the tool above — JPG, PNG, WebP, GIF, HEIC and AVIF are supported.
- Open “AI & provenance” in the report to see whether a C2PA manifest is present, which app signed it and which actions it records.
- Click “Remove AI metadata” and download the cleaned files one by one or as a ZIP. The result panel confirms that C2PA is gone.
See also: Batch remove metadata without uploading to a server
Where the C2PA manifest is stored
The manifest lives in its own block, separate from the image data, so it can be removed without re-encoding:
| Format | C2PA block | Also removed |
|---|---|---|
| JPEG | APP11 (JUMBF) | EXIF, XMP, IPTC, comments |
| PNG | caBX chunk | Text chunks, eXIf, tIME |
| WebP | C2PA chunk | EXIF, XMP |
| HEIC, AVIF | uuid box | Exif and XMP items |
See also: What are C2PA / Content Credentials?
Why people remove Content Credentials
- A real photo was lightly retouched with a generative tool and now gets an “AI info” label it does not deserve.
- The manifest reveals software, account names, edit history or ingredient files you would rather not publish.
- A client or platform needs plain image files without provenance data attached.
- You want to inspect what an image declares before deciding whether to share it as is.
See also: Why are real photos labeled “Made with AI” / “AI info” on Facebook and Instagram? · How to Control Content Credentials in Photoshop and Lightroom (and Clean Your Exports)
What removing C2PA does not do
Stripping the manifest only removes the record stored in that copy of the file. It does not remove invisible pixel watermarks such as SynthID, and it does not affect “durable” Content Credentials that a verifier can look up from a fingerprint or an online repository.
If an image is AI-generated, platforms may still ask you to disclose it. Follow each platform’s rules — removing metadata does not change how the image was made.
See also: SynthID and Invisible AI Watermarks Explained: What Metadata Cleaning Can and Can’t Do · Which Platforms Label AI Images? Instagram, Facebook, TikTok, YouTube and More
Stop new files from getting Content Credentials
In Photoshop and Lightroom, Content Credentials can be turned off in the export settings, although some generative features add them regardless. AI image services usually add them to every output. For files you already have, cleaning a copy here is the quickest fix — keep the original if you may need to prove provenance later.
See also: How to Control Content Credentials in Photoshop and Lightroom (and Clean Your Exports)
Questions about removing C2PA
Does removing C2PA reduce image quality?
No. The manifest is a separate block; the compressed image data is copied unchanged.
Can I see the Content Credentials before removing them?
Yes. The “AI & provenance” panel shows the claim generator, recorded actions and source type, and “All metadata” lists every field.
Are my images uploaded?
No. Files are read and cleaned in your browser and never leave your device.
Will Instagram or Facebook still label the image?
Labels based on C2PA or IPTC metadata lose their source, but platforms can use other signals or ask you to disclose AI content yourself.
Does it remove SynthID or other invisible watermarks?
No. Those are embedded in the pixels, and this tool never changes pixels.
Can I remove C2PA from many images at once?
Yes. Drop up to 50 images per batch and download them all as one ZIP file.