AI Image Metadata: What ChatGPT and Gemini Leave in a File
We downloaded images from ChatGPT and Gemini and read their metadata: a signed C2PA manifest and a declared invisible watermark. What each one contains.
Guides
Guides on metadata, privacy and AI-image provenance, each based on a real file we read field by field.
We downloaded images from ChatGPT and Gemini and read their metadata: a signed C2PA manifest and a declared invisible watermark. What each one contains.
Gemini images carry three marks: a visible sparkle, a C2PA record and an invisible SynthID watermark. Which ones you control, and which ones stay.
SynthID is the invisible watermark Google adds to content made with its AI. How it works, how to check an image for it, and why metadata tools cannot see it.
Your gallery shows a handful of fields. We read a Galaxy S24 photo field by field and found 209, including a country code no gallery displays.
Fix a capture date, add a copyright line, correct a scanned photo. Which of the three date fields to edit, and why writing to one standard is not enough.
Windows, macOS, iPhone, Android and online, with what each method actually leaves behind. Most built-in tools clear far less than people assume.
See a photo's date, camera, settings and location on Windows, macOS, iPhone or in a browser, then work out which of those fields you can actually trust.
Three kinds of evidence say an image was generated, and they are not equally strong. How to read a detector score, and the mistake almost everyone makes.
C2PA attaches a signed record of where an image came from. What that record proves, where it sits in the file, and the three things it cannot tell you.
EXIF, IPTC and XMP all live inside the same photo and do different jobs. What each one stores, where it sits in the file, and how to read or remove it.
iOS only strips GPS from your photos. See exactly what metadata stays on an iPhone image, how to remove location natively, and how to clear the rest.