ScrubPic

Remove AI label data from photos and videos

ScrubPic removes the hidden data (Content Credentials and IPTC AI tags) that Instagram, Facebook, Threads and TikTok read to add AI labels. Invisible watermarks stay, so a label can still appear, and platform rules on labelling AI content still apply.

Free. Your files never leave your device.

How platforms decide to show an AI label

Instagram, Facebook, Threads, TikTok, YouTube and LinkedIn use up to three signals:

Hidden data in the file
Content Credentials (C2PA) and an IPTC tag that says the image was made with AI. Tools such as ChatGPT, Adobe Firefly and Photoshop, and Google’s and Microsoft’s AI image tools add them. Meta and TikTok read this data when you upload.
Invisible watermarks
A pattern hidden in the picture itself. Meta AI adds one to its images, Google’s SynthID is used in images from Google and OpenAI (including ChatGPT), and TikTok is adding its own. These are part of the pixels, not the file’s hidden data.
What you say
The “AI” label you can turn on when posting.

What happens when you remove the hidden data

PlatformWhat it readsAfter ScrubPic removes the hidden data
Instagram, Facebook, ThreadsContent Credentials, IPTC AI tag, invisible watermarksThe “AI info” label that came from the hidden data is no longer triggered. An invisible watermark can still trigger it.
TikTokContent Credentials, its own detection, its own watermarkThe automatic label from Content Credentials is no longer triggered. TikTok’s own detection still applies.
YouTubeContent Credentials, its own detectionThe label that came from Content Credentials is no longer triggered.
LinkedInContent CredentialsThe Content Credentials “CR” icon no longer appears.

How to do it

  1. Add your image in the box above. Nothing is uploaded.
  2. Check what it carries. “This file reveals” shows if it is labelled AI-generated or edited with AI, and which tool made it.
  3. Save the cleaned copy. Content Credentials and AI tags are removed by default, along with location and other hidden info.
  4. Post the cleaned copy, not the original.

What ScrubPic can’t remove

  • Invisible watermarks. They are part of the picture, and ScrubPic never changes the picture.
  • Visible marks, such as a logo or sparkle an AI app draws on the image.
  • Labels added by the app you post from. Images made with an app’s own AI tools, like Meta AI or TikTok’s AI effects, are labelled by that app.

The rules still apply

Removing hidden data doesn’t change what you have to disclose. Meta asks you to label photorealistic video and realistic audio made or changed with AI. TikTok asks you to label realistic AI-generated content, and YouTube asks you to disclose realistic altered or synthetic content. Platforms can add labels themselves or penalise posts that should have been labelled, and in the EU, deepfakes must be disclosed by law from August 2026.

When removing it makes sense

  • A real photo got an AI label because you used a small AI edit, such as removing a stray object with generative remove.
  • Content Credentials can include your name and links to your accounts, which you may not want to share.
  • You want to share an image without the full editing history inside it.

Questions

Will this always remove the “AI info” label on Instagram?

No. It removes the hidden data that platforms read. If the image also carries an invisible watermark, or the platform’s own systems recognise it as AI, the label can still appear.

Does removing the AI info change my picture?

No. ScrubPic only removes the hidden data around the picture. The pixels stay exactly the same, with no re-saving or quality loss.

Why did my real photo get an AI label?

Editing apps add Content Credentials or an AI tag when you use generative AI features, such as generative fill or generative remove. Platforms read that data and may show “AI info”, even for a real photo with a small AI edit.

Do I still have to label AI content myself?

Yes, where the platform’s rules require it. Removing hidden data doesn’t change those rules: Meta, TikTok and YouTube ask you to label realistic AI content, and can penalise posts that should have been labelled.

Related guides