Hoskinson: Midnight Could Privately Authenticate iPhone Photos

Apple’s new Reference Image feature will use signed sensor data to create an unalterable reference photo on the iPhone 18 Pro. Responding to the launch, Charles Hoskinson proposed using Midnight to verify the source through a zero-knowledge proof while keeping the individual device private.

Apple Reference Image Authenticates Photos

Apple introduced Reference Image with the iPhone 18 Pro and iPhone 18 Pro Max. A new sensor in the main camera digitally signs the data recorded during capture.

When Reference mode is enabled, Apple’s Private Cloud Compute processes the signed sensor data into an unalterable reference image. The result appears alongside the main photo in the Photos app and functions as a digital negative that can be compared with edited versions.

Apple is also providing APIs across iOS, iPadOS and macOS, allowing third-party applications to display and inspect authenticated images. The system is intended for photographers, photojournalists and other users who need evidence of what a camera recorded as AI-generated and manipulated media becomes harder to identify.

Hoskinson Adds a Zero-Knowledge Model

Responding to Apple’s announcement, Hoskinson wrote, “Midnight fixes this.” He proposed a zero-knowledge proof confirming that a photograph came from a genuine Apple phone without revealing which device produced it.

His model also includes selective disclosure, allowing users to reveal only the information required for verification. A photograph could therefore carry proof of an authentic hardware source without exposing additional data connected to the phone or its owner.

Midnight provides programmable privacy through zero-knowledge technology. Applications can verify selected facts while keeping the underlying information private. Its token NIGHT generates DUST, the resource used to process transactions on the network.

A Concept Without a Midnight Integration

Hoskinson has not published a technical specification or prototype connecting Apple Reference Image with Midnight. Apple has also not announced a partnership or integration with the network.

The proposal defines a potential extension of Apple’s authentication system. Apple can confirm what its camera sensor recorded and provide a reference for detecting later edits. A Midnight layer could convert that confirmation into a proof that third parties can verify without identifying the individual device.

For the wider Cardano (ADA) ecosystem, the idea connects programmable privacy with the growing need to distinguish authentic photography from manipulated or synthetic media. Turning it into a product would require a technical method for converting Apple’s signed sensor data into an independently verifiable zero-knowledge proof.