Apple is working on Apple Reference Image, a photo provenance system that can verify whether an image was captured with an iPhone. References to the feature appear in a privacy disclosure included in iOS 27 beta 5, though the feature is not yet live.
The Camera app will include a Reference mode users can opt into; images taken in that mode will embed provenance data in the file’s metadata.
How Apple Reference Image authentication works
Tapping a Reference badge on an image initiates authentication: the raw image, sensor signatures, the capture time frame and the sensor’s unique hardware identifiers are sent to Apple’s Private Cloud Compute (PCC) servers. PCC uses that information to determine whether the camera captured the photo, assigns it a unique ID, and returns an authenticated version to the user’s device.
When a photo is authenticated, Apple receives sensor data, a hash and the assessment results, but not the raw photo. Apple can refuse to authenticate images from a sensor that it considers potentially compromised and can retroactively revoke a sensor’s past authentications.
Authenticated images can be viewed on iPhone, iPad or Mac, and the Reference badge is clickable on macOS. Because images must be captured in Apple’s specific Reference mode to be authenticated, the feature appears aimed at professionals such as pro photographers and journalists who need verifiable provenance. The disclosure also mentions “photos or videos” and “uncropped footage,” indicating a possible video component.
Reference mode would provide hardware-backed evidence that an image came from a physical camera sensor at a time when AI can generate realistic-looking images. AI-generated images are often watermarked or labeled via metadata to mark them as synthetic, including images from Image Playground.
Camera manufacturers such as Leica, Sony and Nikon use C2PA Content Credentials for authentication, and Google has adopted C2PA for the Pixel 10 lineup. Apple’s system appears intended to serve a similar purpose.
