Will Instagram label my photo as AI?
Instagram and Facebook decide at upload, and they read hidden tags inside the file, not just the picture. When an editing app records that generative AI made or changed the image, Meta shows the label “AI content” (formerly “AI info”). TikTok shows “AI-generated”, Pinterest shows “AI modified”, and LinkedIn shows a small “CR” icon for Content Credentials.
That is why a real photo can get labeled. One Generative Remove pass on a stray cable, an AI denoise or a phone eraser can tag the whole export. The label describes the file, not your camera. Meta says it reads industry-standard indicators such as C2PA and IPTC, so reading the metadata answers the most common version of the question before you post.
What does this AI label checker read?
Drop the file above and the checker opens it. It lists what it finds in plain words, together with the app that wrote each entry and where in the file it sits:
- C2PA Content Credentials: the signed record of who made the file, with which app, which edits were recorded and whether it declares generative AI.
- IPTC Digital Source Type: one field that can say “trainedAlgorithmicMedia” (made by generative AI) or “compositeWithTrainedAlgorithmicMedia” (a real image edited with generative AI).
- XMP AI fields: the AI system used, prompt details, and editing-history steps that name an AI tool.
- Generator traces: EXIF Software tags, comments and PNG text chunks with prompts or settings from tools such as Stable Diffusion and ComfyUI.
- Google AI provenance hints: metadata that names a Google AI tool, such as Gemini, Imagen, Veo or a Google Photos AI edit.
- Private details: GPS location, camera serial number and owner name, so you can decide what to share.
How to read the result
The verdict has three levels. “Likely to get an AI content label” means the file declares AI generation or AI editing in a field that platforms read, usually C2PA or the Digital Source Type. “Some apps may label this” means weaker hints, such as the name of an AI tool in a text field. “No AI tags found” means the metadata is clean.
Read the wording too. A file that says “edited with AI” is not the same as a file that says “created by AI”. If you took the photo and only used AI to remove a distraction, the label would be a false positive, and the result shows exactly which tag causes it.
Which apps write AI tags into real photos?
Most false labels start with an everyday edit. These apps can write AI tags into a file even after a small change:
- Adobe Photoshop and Lightroom: Generative Fill, Generative Remove and AI Denoise
- Adobe Firefly and Adobe Express
- Canva Magic tools
- Google Photos Magic Editor and AI edits on Pixel phones
- Samsung Galaxy AI edits
- Apple Clean Up and Image Playground
- ChatGPT and other OpenAI image tools, Microsoft Designer and Copilot, and Midjourney
SynthID and other pixel watermarks — handled too
Some AI tools also hide a watermark inside the pixels. Google SynthID is one example: Google adds it to images and videos from its generative AI tools, and there is no switch to turn it off. This kind of mark is spread across the image itself rather than stored in a metadata field.
We handle it. When the file’s metadata names a Google AI tool, the result flags that SynthID is present, and cleaning removes it along with the tags, so the clean copy carries no AI marks for any checker to find.
A file check still cannot change a platform’s own AI classifier, the AI label toggle you switch on when posting, or Instagram’s account-level “AI-generated profile” label, which is set on the account and not read from a file.
What to do with each result
The check is only useful if it tells you what to do next. Here is how each outcome plays out:
- AI tags on a real photo you edited: tap “Delete AI label” and post the clean copy. The default mode keeps camera details, location and the color profile.
- AI tags on an image that AI generated: the tags are accurate. Keep the AI disclosure when you post.
- A Google AI tool named in the file: tap “Delete AI label”, and cleaning removes both the metadata tags and the SynthID pixel watermark in one pass.
- No AI tags, but a post was labeled anyway: the cause is outside the metadata, such as your own AI label setting or the platform’s own judgment.