AI label guide · November 2026

Why does my photo say “AI info”?

You took the photo. You were there. And your post says AI info. Nobody at Instagram or Facebook looked at your picture and doubted it: an app that handled the file left a note inside it, and the platform read the note. Here’s how to find out which app, and how to post without it.

Download for MacmacOS 14+ · Apple SiliconFree trial Download for WindowsWindows 10 & 11 · Microsoft StoreFree trial
Watch how it works 1:24

Free for 3 days · then $5 once · no subscription · Buy a licence

IMG_2041.jpg2 findings
  • Content CredentialsSigned after an object was erasedFoundRemoved
  • AI source typeComposite with trained algorithmic mediaFoundRemoved
  • The photo itselfTaken by you · pixels unchangedFoundKept
Clean copy verified · pixels unchanged

Short answer

Your photo says “AI info” (Meta’s help pages now call it the AI content label) because its file carries AI metadata:

  1. An app wrote it. Usually after one AI-assisted edit: an erased object, an extended background, a generated fill.
  2. The platform read it when you uploaded, and labelled the post.
  3. You can see it with the free checker below, and remove it from a copy of the file with MetaClean before you post.

Get MetaClean freeFree for 3 days · then $5 once · no subscription

Watch: why a real photo gets “AI info”, and how to remove the note in the file before you post (1:24, captions on screen).

The reason, in one line

The label describes your file, not your photo. Instagram’s help page puts it this way: content made or modified with AI tools may carry an industry-standard signal, and Meta’s systems read those signals to decide whether the content needs a label. Meta has said which signals it means, the AI information in the C2PA and IPTC standards.

Both are metadata: text stored in the file beside the picture. Your camera didn’t put it there. An app did, later.

Check your photo for the AI note

Use the exact file you uploaded. The checker reads it on your device, so nothing is sent anywhere.

If it finds Content Credentials or an AI source tag, you have your answer, and often the name of the AI tool that was used. If it finds nothing, skip to when it isn’t the file.

Which app added it

These are the usual sources. In each case the photo or video can be entirely real apart from one small edit:

AppWhat triggers itWhat it writes
PhotoshopGenerative Fill, Generative Expand, RemoveContent Credentials, attached automatically
LightroomContent Credentials applied at exportContent Credentials
Premiere ProGenerative ExtendContent Credentials in the MP4 or MOV
CapCutAI effects and templatesAI markers in the exported video
Your phone’s editorObject eraser, AI expand, reimagineAn AI source tag or Content Credentials
An AI generatorThe whole imageContent Credentials, often with an invisible watermark too

Two things catch people out. The note survives saving, renaming and sending the file to yourself. And it travels: a photo your retoucher touched arrives already marked, even though you never opened an AI tool.

You removed a cable.The app wrote “AI”.The platform believed the app.

The picture is real. The file just has one line too many.

On the post, or in the menu?

Where the label sits tells you what the file claims. Meta’s current help pages draw the line like this:

In the post’s Options menu

Content whose signals say it was only modified with AI doesn’t get a label directly on the post. The information is under the Options menu of the post, story, reel or thread.

On the post itself

The signals in the file say AI created it, or Meta’s systems detected the image as AI-generated. Some editors mark a lightly edited photo this strongly.

Either way, a client, an editor or a competitor who taps through sees “AI” beside your work. The same rule applies on Threads. Platform by platform: Instagram, Facebook, TikTok, YouTube, Pinterest and LinkedIn.

How to stop your photos saying “AI info”

  1. 1

    Open the file in MetaClean

    On your Mac or Windows PC. It lists the AI markers it finds, in plain words.

  2. 2

    Choose Credentials only

    The Content Credentials and AI tags are removed from a copy. Everything else, and every pixel, stays.

  3. 3

    Post the clean_ copy

    For a post that’s already labelled, delete it and share the clean copy instead.

  4. 4

    Catch the next one automatically

    Add your export folder to Watch Folders and MetaClean flags any new file that carries an AI note, before you post it.

People often screenshot the photo instead. It works, and it costs most of the resolution: see does a screenshot remove AI metadata?

When it isn’t the file

If the checker finds no AI metadata, one of these is the reason, and no metadata tool changes it:

  • You or the app labelled it. The “Add AI label” switch was on when you shared, or you posted straight from a Meta AI tool.
  • Meta’s systems detected it. Meta says images still receive a label when its systems detect they were AI-generated.
  • An invisible watermark. Some generators mark the pixels as well as the metadata. That stays with the picture.

For a real photo with one AI touch-up, these rarely apply. The note in the file is the reason nearly every time, which is why checking the file comes first.

Your photos stay on your computer

MetaClean works on your own Mac or Windows PC: no uploads, no account, no file size limits. More about privacy →

People are asking

It’s a real photo

Why, if it’s not AI?How does Instagram know?Which app did it?On the post or in the menu?How do I check?How do I stop it?

“AI info” on a real photo: your questions

Why does my photo say AI info when it’s not AI?

Because the file says so, not the picture. An app that handled the photo wrote a note into its metadata, usually C2PA Content Credentials or an IPTC “digital source type” tag, and Instagram and Facebook read that note when you upload. A single AI-assisted edit, like erasing an object, is enough for an app to add it.

How does Instagram know my photo was edited with AI?

Instagram’s help page says some content made or modified with AI tools carries an industry-standard signal, and that Meta’s systems read those signals to decide whether the content needs a label. Meta has named the standards: the AI information in C2PA and IPTC metadata. It also says its systems can label images they detect as AI-generated.

Why does it say AI info on my post but not on my friend’s?

Your files went through different apps. The same photo sent straight from the camera has no AI note; the copy that passed through an editor with a generative tool does. Check the exact file you uploaded.

What is the difference between the label on the post and the one in the menu?

Meta says content whose signals show it was only modified with AI doesn’t get a label directly on the post, and the information sits under the post’s Options menu instead. A label on the post itself means the signals in the file say AI created it, or Meta’s systems detected it as AI-generated.

How do I check if my photo has AI metadata?

Drop it into the checker on this page. It reads the file in your browser, without uploading it, and tells you whether it carries Content Credentials or an AI source tag, and names the AI tool when the file does.

How do I stop my photos being labelled AI info?

Remove the AI metadata from the file before you post. MetaClean does that on your Mac or Windows PC: it takes the Content Credentials and AI tags out of a copy, leaves the image untouched and checks the result. It is free for 3 days, then $5 once.

Does the label go away if I edit the post?

Meta’s help pages describe how to add an AI label yourself, not how to take off one its systems applied. What you control is the file: post a copy without the AI metadata.

Your photo is real.
Let the file say so.

MetaClean shows the AI note in your photo and removes it on your own Mac or PC. Free for 3 days, then $5 once.

Download free trial