iOS 27.2 hides Google's SynthID mark in some AI-edited photos, but Apple can't read it

What Apple added

Apple's iOS 27.2 update brings Google's SynthID technology to some photos that AI has created or edited. SynthID is a watermark you cannot see, built into the image itself. As infosecu.technews.tw reported, the goal is to make it easier to tell when a picture has been through AI.

It applies to Apple's own Photos app and its Image Playground tool, which generates images.

When the mark appears

Not every edit gets tagged. Anything made in Image Playground carries SynthID automatically. In Photos, it depends on the feature you use.

The mark is written into the pixels when you use Extend, Spatial Reframing or a bigger Clean Up edit. Because it lives in the pixels, it is hard to strip out unless the image is heavily cropped or regenerated.

The catch: checking it

You cannot check for the watermark inside iOS or any Apple app. You have to upload the file to Google's SynthID detection website.

Google's Gemini can detect marks too, but Google only guarantees it will recognize content made by Google AI. It has not made the same promise for images made on an iPhone.

Privacy and limits

Google says files uploaded to the detector are handled and deleted once the result appears. It still keeps the file's digital signature for 24 hours to block automated abuse and limit use. If you sign in with a third-party account, Google keeps basic account identity details.

So think twice before uploading private or sensitive photos.

SynthID also reads a hidden mark; it does not judge whether a picture looks AI-made. An AI image created without the mark will pass unnoticed, however artificial it looks.