How to Post Nano Banana Images to Instagram Without the "Made with AI" Label

Nano Banana — Google’s image model — turns out wildly good edits and generations, and Instagram is the obvious place for them. Post one, though, and you’ll often see a “Made with AI” label with the “AI info” panel attached. Nano Banana is a more interesting case than most generators, because Google marks its images in a way that’s worth understanding before you try to clean them.

Nano Banana marks images two different ways

Google attaches two distinct things to a Nano Banana image:

1. SynthID — an invisible watermark embedded in the pixels themselves. It’s designed to be robust: it’s meant to survive resizing, compression, screenshots and even re-encoding. This is Google’s own provenance signal.
2. C2PA / metadata credentials — the standard “made with AI” record written into the file’s metadata, the same provenance layer most AI tools use.

These are not the same thing, and the distinction is the whole story here.

Which one does Instagram actually read?

This is where being precise matters. Instagram’s “Made with AI” label is applied from the C2PA content credentials and metadata it parses on upload — the same mechanism that flags Midjourney, DALL·E and the rest. (See Instagram’s “Made with AI” label, explained and what C2PA content credentials are.)

Instagram does not currently apply that label by scanning for Google’s SynthID. So even though SynthID is the more robust watermark, it isn’t the thing triggering the tag on your post — the metadata credential is.

That difference is exactly why cleaning works in practice.

Why a native re-encode clears the label

A native re-encode rebuilds the image from scratch through Apple’s CoreImage pipeline. The C2PA credential and metadata don’t carry into the new file — so the signal Instagram reads is gone, and the “Made with AI” label has nothing to apply itself from. In real-world testing, cleaned Nano Banana images post to Instagram without the tag.

Here’s the honest boundary, stated plainly: a re-encode removes the metadata/C2PA that drives the label. We are not claiming it strips SynthID — that watermark is pixel-level and built to persist, and Google’s own detection tools could still recognize it. But since Instagram’s label keys off the metadata rather than SynthID, clearing the metadata is what removes the label you actually see. We’d rather be exact about that than wave our hands.

The workflow

1. Download your Nano Banana image at full resolution (don’t screenshot it).
2. Get it onto your iPhone so it’s in Photos.
3. Open CleanAi and pick the image — processed on-device, never uploaded.
4. Crop if you like to tidy the framing.
5. Tap Clean to re-encode through CoreImage and drop the C2PA credential + metadata.
6. Post to Instagram — feed, carousel, Reel cover or Story.

Quality is preserved

CleanAi exports with native iOS presets, so the cleaned copy keeps its resolution with no visible compression added. A real re-encode, not a lossy screenshot — your Nano Banana render looks identical, just without the metadata credential attached.

What doesn’t remove the label

• Cropping alone changes pixels but leaves the C2PA metadata in place.
• Screenshotting degrades quality and isn’t a reliable metadata wipe.
• Re-saving the same file typically carries the credential straight through.

More background: why Instagram says your photo is AI.

Scope

This is for your own Nano Banana generations. CleanAi is on-device only, never connects to Instagram, and never touches anyone else’s content — you’re responsible for posting work you have the right to share.

Posting the same images to TikTok too? See the Nano Banana → TikTok companion guide, or jump to the full step-by-step for the “Made with AI” label on Instagram.