The other day, I was on the phone with a photographer who I’ve been friends with for a decade. We were discussing the recent photograph of Donald Trump being shot — and she related to me that her roommate didn’t think it was real. She knew better, but she also understands that we live in a world where the image could’ve been forged and made by AI or Photoshop. Indeed, one of the hottest topics in the world of digital photography these days has to do with content authentication. While many brands try to do it via firmware in-camera, some work to put dedicated chips in their cameras.
Specifically, one of the brands that has been huge on using a chip is Leica. Leica is one of the most celebrated brands in photography authenticity. Their cameras have been used to capture some of history’s most important moments. When our workflows were primarily film-based, we didn’t have many problems with photography manipulation. But in the digital world, society’s reliance on social media has made it a much more serious concern.
Most recently, Leica has been working to integrate Content Authenticity into its cameras. The technology works alongside Adobe’s Content Authenticity Initiative, which the Phoblographer is a member of.
“Content Credentials are exclusive to the Leica M11-P,” says Nico Köhler, Head of Product Experience at Leica Camera AG in an interview with the Phoblographer. “The complexity of new technology development necessitates prioritization and resource allocation. Implementing this function required market readiness and technical advancements.” He continues to state that Leica hasn’t been able to do with the Leica SL3 — the camera that arguably more photojournalists might use. According to Nico, it would’ve posed a challenge that delayed the product launch.
The Leica M11-P uses a dedicated secure chipset in the camera to protect content credentials. Specifically, it includes private keys that work with the C2PA-compliant algorithm, which verifies the history of the edits in the image and the fact that someone actually shot the image. According to Leica, security is the primary difference between hardware and software solutions.
Japanese manufacturers, like Sony, have been looking at being C2PA compliant through the software route. According to their press release, this feature is only available to the Alpha 1, Alpha 9 III, Alpha 7S III, and Alpha 7 IV. There’s sadly no plan to put it into the Sony a7c II or the Sony a7c R. Considering Sony’s history, even if it were done via a firmware update, they’d make you buy a new camera. It’s a tactic that has backfired on other camera makers.
However, this means that if you’ve spent thousands of dollars on older Leica cameras (or cameras in general), then you might be out of luck. While Köhler says that adding Content Credentials via the Leica FOTOS app is feasible, it’s not as secure of an option. “Delivering private keys via the app could compromise security. Leica FOTOS app currently only displays if photos are signed using C2PA Rust SDK, demonstrating its capability.” He continues to state that adding a device into the hot shoe or a port is challenging — especially as it requires core camera integration.
From Nico’s words, we can assume that it’s as core as implementing certain features into the menus.
Several photographers aren’t necessarily aware of how important this is, but the recent photograph of Trump being shot at is a great example. If you’re not a photojournalist, though, you still might not care until you realize that the photos you’ve uploaded to various services are being used to feed AI. So why should other photographers care about content credentials?
“Content Credentials provide proof that a photo was captured with a Leica camera at a specific time and location, adding value and authenticity to their work,” Nico explains about Leica’s devices. “This protection is crucial amid the rise of generative AI-generated content, ensuring trust and integrity in photography.”
This leaves us asking: if software can manipulate the makeup of an image, is it really more secure? How different is it from something like Adobe’s Intelligent Upscaling? We know from testing it that the images aren’t anywhere as good as when they come from a high-resolution camera in the first place. So is this the same thing?