How to tell if an image was made with Gemini

What Gemini writes into the file

This applies to the image models Google calls “Nano Banana”, in the Gemini app, the Gemini API and Vertex AI. Imagen models are missing from Google's list of models with Content Credentials.

C2PA (Content Credentials) is a signed provenance record in the file, IPTC Digital Source Type a metadata field on how the image was made, for example trainedAlgorithmicMedia for fully AI-generated images. An invisible watermark sits in the pixels, and only the vendor can read it.

How to check a Gemini image

1. Drag the original file from Gemini into the box above. It is read in your browser, not uploaded. Screenshots and copies from stores or social networks usually no longer carry the data.
2. Under “Content Credentials (C2PA)”, the “Certificate (organisation)” row then names Google LLC. According to Google, the record identifies the image as AI-generated.
3. Under “IPTC / XMP” you may also see the Digital Source Type trainedAlgorithmicMedia.
4. The Gemini app checks SynthID. Upload the image there and ask whether it was made with Google AI.

More detail in the Content Credentials viewer and AI image checker.

The checker is not an AI detector. It only shows what the file says. An empty result does not prove a real photo, not least because providers of AI systems placed on the market before 2 August 2026 have until 2 December 2026 to add machine-readable marking (Article 111(4) AI Act).

What survives uploads and edits

Google describes SynthID as resistant to cropping, filters and lossy compression. The C2PA data, by contrast, is metadata and is lost through screenshots and many uploads. If a program without C2PA support edits the file, Google treats that as tampering, and validation fails.

The visible sparkle sits in the image itself, but it is missing on images from Google AI Ultra and Google AI Studio.

Labelling duty in your store under the EU AI Act

If you use Gemini to create or edit images for your own store, you are a deployer, even when an agency or an employee works on your instructions (Article 3(4) AI Act, Commission guidelines para. 12, 14). If the image is a deep fake, meaning realistic and falsely appearing authentic, you have had to label it visibly since 2 August 2026 (Article 50(4) AI Act). The labelling check tells you whether that applies, and the guide to the labelling duty explains the rules.

The traces in the table do not replace that label, because people cannot see them straight away (guidelines para. 117). Still, keep them in the file, as the guidelines recommend. They document where the image came from.

Gemini creates photorealistic scenes and edits uploaded product photos, and both can be a deep fake. The sparkle does not say what the AI did.

How Gemini labels its own output

Besides SynthID, C2PA and the sparkle, Google offers the check in the Gemini app. In November 2025 Google announced that it would extend it to C2PA records from other providers.