Technology
Apple developing system to authenticate iPhone photos as genuine
A forthcoming iOS feature could let users prove their iPhone images are authentic by embedding verifiable metadata at the moment of capture.
The Verge
Apple is working on a verification system that would allow iPhone owners to authenticate their photographs and demonstrate they weren't created or manipulated with AI tools. Code discovered in iOS 27 beta 5 suggests the company is building an "Apple Reference Image" feature that embeds provenance data directly into photos as they're shot.
How the system works#
The authentication process would begin when users enable the feature through Settings > Camera > Reference Image > Reference Mode. Only photos captured using a dedicated "Reference" mode would carry the necessary verification markers. To authenticate a particular image, users would tap a Reference badge on the photo, triggering a verification request to Apple's Private Cloud Compute servers.
The authentication process sends the raw image along with embedded metadata—including sensor signatures, capture timestamp information, and unique hardware identifiers—to Apple's servers for verification. The system can then return an authenticated version assigned a unique identifier. Importantly, Apple does not retain the actual photo during this process, though it may collect sensor data to identify and block images from compromised cameras or revoke authentication on previously verified photos if their source hardware is found to be problematic.
Why Apple's approach differs#
This authentication system operates similarly to the C2PA Content Credentials standard, a provenance framework that tracks image origins and AI involvement. Camera manufacturers including Canon, Nikon, Sony, FujiFilm, and Leica have gradually integrated C2PA support into their equipment, and Google's Pixel 10 phones also support it. Apple has notably resisted adopting this existing standard.
The company may be betting that its proprietary system will prove more effective and reliable than C2PA, which has faced scrutiny over inconsistent detection capabilities. Building its own solution also aligns with Apple's broader push to help platforms label and authenticate human-made content—a goal some industry leaders believe is more achievable than attempting to identify all AI-generated material.
Current status#
The Apple Reference Image feature is not yet active in iOS and will be disabled by default when released. A privacy disclosure in the beta notes indicates users can manually activate the system. Authenticated images can be viewed across iPhone, iPad, and Mac devices once the feature launches.

