Touched up a real photo with Generative Fill and got tagged as AI? Photoshop writes C2PA Content Credentials into the file, and platforms read them. StripShot removes that metadata at the binary level in your browser, with zero quality loss, so your real photo stops getting mislabeled.
Last updated June 2026
This is the most common false positive on the internet right now. You shoot a real photograph, remove a distraction with Generative Fill or extend the frame with Generative Expand, and export. Photoshop writes C2PA Content Credentials declaring AI involvement into the file. When you upload it, Instagram, Facebook, LinkedIn, or Pinterest reads that tag and stamps "Made with AI" on a photo that is 99% real.
The label is driven by the metadata, not by an analysis of your pixels. Remove the metadata and the automatic trigger is gone.
Triggers the AI label
Does not trigger it
Export from Photoshop
Export or Save As a JPEG, PNG, or WebP. Use the original export, not a screenshot, to keep full quality.
Drop it into StripShot
Drag the file into the tool. It detects the C2PA Content Credentials and Adobe XMP tags Photoshop embedded.
Strip the AI metadata
Click strip. The C2PA and XMP provenance data is removed at the binary level while your pixels stay untouched.
Re-upload the clean file
Post the cleaned copy. With the metadata trigger gone, the platform has nothing to read for the label.
The instinct is to re-save the file in Photoshop with metadata turned off, but that recompresses your carefully edited pixels. StripShot removes only the C2PA and XMP marker blocks and writes your original pixels straight back. Your retouching is preserved exactly; only the AI tag is gone.
If you used a generative feature, your image does contain some AI-generated pixels. Removing the metadata stops automatic mislabeling, but how you disclose AI use when you publish is your call and may be governed by platform terms and, from August 2026, the EU AI Act. StripShot controls your file's metadata; it does not make that decision for you. See our disclaimer.
Photoshop writes C2PA Content Credentials into your file whenever you use a generative AI feature such as Generative Fill or Generative Expand, even if you only touched a small area of a real photograph. Instagram and other platforms read that C2PA data and apply the 'Made with AI' label automatically. The label reflects the metadata, not how much of the image is actually AI.
Generative Fill, Generative Expand, Generative Remove, and the Firefly-powered generation features all embed C2PA Content Credentials. Standard edits like crop, levels, healing brush, and manual cloning do not.
Export your image, drop the JPEG, PNG, or WebP into StripShot, and click strip. It removes the C2PA Content Credentials and Adobe XMP tags at the binary level, then you re-upload the clean file. Your pixels are never re-encoded.
No. StripShot removes only the metadata marker blocks and writes your original pixel bytes back unchanged. Re-saving in Photoshop to drop metadata, by contrast, recompresses the image.
If you used a generative AI feature, your image does contain AI-generated pixels, and how you disclose that when publishing is your responsibility and may be governed by platform terms and the EU AI Act. StripShot gives you control over your file's metadata; it does not make that disclosure decision for you. See our disclaimer.
StripShot removes the C2PA and XMP metadata Firefly and Photoshop embed. It does not alter pixel-level watermarks. For images generated entirely in Firefly, see our Firefly guide.
The full guide to the metadata behind the label.
For images generated fully in Firefly.
Every generator, C2PA, and XMP tag in one tool.
See the C2PA tags Photoshop wrote before you strip them.
Pricing
No card needed to start. Upgrade anytime.