Paranormal Unsettled

The paranormal world is already starting to face a problem unlike anything it has dealt with before.

AI can now create or manipulate photographs or videos with incredible realism. A strange figure in a hallway, a face in a window, a shadowy apparition, even an entire scene can potentially be generated or altered without leaving obvious signs that anything was changed.

Photographs have always been some of the most recognizable forms of paranormal evidence. But if anyone can create a convincing "ghost photograph" with AI, how do we know what we're looking at is actually something a camera captured?

Apple may have an interesting potential answer.

What If Your Camera Could Prove What It Captured?

Apple has introduced a technology called "Apple Reference Image", debuting on the main camera of the iPhone 18 Pro and iPhone 18 Pro Max.

It's an opt-in camera mode designed to create a cryptographically secured reference image showing what the iPhone's camera sensor actually captured.

The important part is that the protection begins at the camera sensor itself. When a photograph is taken in Reference mode, the sensor cryptographically signs the captured pixel data immediately. This creates what Apple calls a "secure digital negative," containing the image data, important sensor information, and secure timestamp information.

That could become extremely interesting for paranormal evidence, if an iPhone is being used, which is becoming more and more common as the evolution of digital video and content creation advances. 

This also helps investigators that DO capture real evidence, but perhaps in this day and age is called out for using AI, even when they are not.

Everyone else has to take their word for it but with a verified reference image, there could potentially be another layer of evidence, the phone can establish that the image originated from its physical camera sensor.

What About AI Manipulation?

This is where things get really interesting.

Apple's system is designed so that the original sensor data is cryptographically protected before the operating system can manipulate it. The resulting image is then developed through Apple's Private Cloud Compute environment, where Apple says the processing can be independently verified.

That means the system isn't simply saying: "Trust us, this photo is real." It's creating a chain of cryptographic evidence beginning with the physical camera sensor.

If someone simply creates an image with an AI generator, there is no original iPhone sensor capture behind it. If someone takes a normal photograph and later uses AI to add a shadow or an apparition, that alteration would not change the fact that the original Reference Image capture contained something different.

The crucial question becomes, can the final image be cryptographically traced back to what the sensor actually captured? That's a very different standard from simply examining an image and trying to spot AI artifacts.

What If Someone Uses AI Afterward?

This is where the future gets particularly interesting. Suppose an investigator captures a legitimate Reference Image showing an unexplained object. Later, somebody downloads that image and uses AI to add a giant ghost behind it.

The manipulated version might look completely convincing. But the AI generated version wouldn't magically acquire the cryptographic history of the original camera capture. The original authenticated image would still be the reference point.

That could make it possible to distinguish between: "This is what the iPhone actually captured." and "This is what somebody created afterward." Apple also designed the system with revocation capabilities. If a sensor were eventually determined to have been compromised, Apple says individual images or images associated with a particular sensor could be revoked.

This Could Go Beyond Paranormal Photography

The same problem exists with video. AI generated and manipulated video is rapidly becoming more convincing. Eventually, determining whether something actually happened in front of a camera could become extremely difficult simply by looking at the footage. 

A future version of this type of technology could potentially provide a similar foundation for video captured on an iPhone. The ideology behind it would be the same.

Apple's current announcement is specifically about reference photographs, so we shouldn't assume that it already provides this protection for video. But the underlying concept could eventually have applications far beyond still images.

And if technologies like Apple Reference Image become widespread, an investigator using an iPhone could eventually have something that paranormal investigators have never really had before, a way to establish a  protected connection between the photograph being presented and the physical camera sensor that captured it.

It still won't tell us whether the figure in the hallway is ACTUALLY a ghost, but it can ask, "did the camera actually capture this, or did someone create it afterward?"

In a world increasingly flooded with AI generated "paranormal evidence," that distinction could become incredibly important.