According to a report by MacRumors yesterday, Apple has quietly integrated Google's SynthID technology in iOS 27.2. From now on, images generated with "Image Playground" and some photos edited with Apple Intelligence will be stamped with a hidden watermark invisible to the naked eye.
SynthID is a generative AI content identification solution built by Google, embedding markers directly at the pixel level, specifically to "stamp" AI images and videos. Google says it neither damages image quality nor can it withstand cropping, filters, and compression and other post-processing.
In hands-on testing, using the "Expand" and "Reframe" editing features in the "Photos" app resulted in detection as AI-generated; whether the "Clean Up" feature gets watermarked depends on the extent of the changes—minor retouching carries no watermark, but larger changes can't escape it. And any image produced by Image Playground comes with SynthID built in.
To verify authenticity, you have to go to a Google website
One awkward point is that the iPhone itself provides no detection entry point; users must go to the SynthID Detector web page to verify whether an image carries an AI watermark. This move echoes the global call for mandatory labeling of AI-generated content, and gives ordinary users an extra card to check when they come across suspicious images while scrolling. As phone makers make "provenance labeling" a system-level default action, the line of defense for distinguishing real from fake content is slowly being drawn from third-party tools into the operating system itself.