iPhone Now Embeds Google's SynthID Watermark in AI Images and Edits
Users cannot check for the mark on the phone itself and must upload images to a separate website to detect it.
Apple has begun embedding Google's SynthID watermark into AI-generated and AI-edited images on iPhone, with the change now active in iOS 27.2 after Apple first flagged it in June. The invisible marking sits at the pixel level rather than in metadata, which Google says leaves image quality unchanged and lets the mark survive cropping, filtering and compression. Every image made in Image Playground carries it, while edits made with Apple Intelligence in Photos vary: Extend and Reframe results were marked in testing, and Clean Up added the mark only when the edit was large. iPhone users have no built-in way to detect the watermark, so checking an image means uploading it to the SynthID Detector website.
Why it matters
The watermark is invisible and sits in the pixels rather than in metadata, so it is meant to persist through cropping, filters and compression. That makes it harder for anyone to strip out by simple editing. Coverage is not uniform, though: every Image Playground creation is marked, while Photos edits vary, with Clean Up marking only larger changes. Anyone relying on the mark to judge whether an image is AI-made should keep that gap in mind.
Matters or hype?
Does this story matter, or is it hype? Vote before you see what everyone else said.