Fix the "AI info" label on your Instagram photos

Instagram labels a photo as AI when the file says an AI tool touched it, even if you only removed a stray object. Check your photo here and remove the markers before you post.

JPEG, PNG and WebP are cleaned on your device and never uploaded.

Edits that commonly trigger the label
PS
Photoshop Generative Fill or Remove tool
LR
Lightroom Generative Remove or AI Denoise
Phone
Magic Eraser, Galaxy AI or Clean Up edits
Apps
AI upscalers, background removers and filters

Why Instagram says "AI info"

Since 2024, Meta reads two industry markers when you upload a photo to Instagram, Facebook or Threads:

  • C2PA Content Credentials: a signed record of how the image was made, added by ChatGPT, Adobe Firefly, Photoshop, Google and Microsoft tools.
  • IPTC digital source type: a tag in the file's XMP data that marks an image as "trained algorithmic media" (AI-generated) or "composite with trained algorithmic media" (edited with AI).

If either is present, the post gets an "AI info" label. The markers don't say how much AI was used, so a real photo with one small generative edit is treated like a fully generated image.

How to post without the label

  1. Clean the original file

    Add the photo above. AI markers are highlighted; select Remove metadata.

  2. Download the clean copy

    The pixels are unchanged, so quality is identical.

  3. Delete the labelled post

    A label added from metadata usually stays on a published post.

  4. Upload the clean copy

    Share the downloaded file, not the original from your camera roll.

Export settings that avoid the problem

You can also stop the markers being written in the first place:

  • Photoshop: use File, Export, Export As, and set Metadata to None. Content Credentials can be turned off under Window, Content Credentials.
  • Lightroom: in the Export dialog, set Metadata to "Copyright Only" and leave Content Credentials off.
  • Phones: sending a photo through a messaging app usually strips metadata, but it also compresses the image. Cleaning here keeps full quality.

What this can't do. Meta can also label images based on its own analysis, which metadata removal doesn't affect. And if you post photorealistic AI-generated video or audio, Meta's rules ask you to disclose it yourself. This tool is meant for real photos labelled by mistake and for protecting your privacy.

Questions

Can I remove the label from a post that is already up?

In most cases, no. A label Instagram adds from file metadata stays on the published post. Delete the post, clean the original photo here and upload it again.

Will a cleaned photo never get the label?

Not from metadata, because the markers Instagram reads are gone. Meta also runs its own checks and can add a label based on the image itself. Removing metadata does not change that.

Is "Made with AI" the same as "AI info"?

Yes. Meta first called the label "Made with AI" and renamed it "AI info" in July 2024 after photographers complained that lightly edited photos were being labelled. It is triggered by the same metadata.

Does this work for Facebook, Threads and LinkedIn?

Yes. Facebook and Threads use the same system as Instagram. LinkedIn shows a "CR" Content Credentials icon from the same C2PA data, which a cleaned file no longer carries.

Do I still need to label AI content myself?

Meta asks you to disclose photorealistic video and realistic-sounding audio made or changed with AI, and some countries have their own rules. Metadata removal is meant for real photos labelled by mistake and for privacy, not for hiding AI content you are required to disclose.

Are my photos uploaded?

JPEG, PNG and WebP files are read and cleaned by your browser and never leave your device. HEIC and TIFF files are sent over an encrypted connection to our server, where ExifTool cleans them. Each upload is deleted as soon as you download the clean copy, and always within 30 minutes. Images are never stored, shared or used for training.