Every Camera Brand Agrees: Generative AI Doesn’t Belong in Photography

By and large, big tech executives love AI. There is some reporting that executives now say “AI” more than they say “earnings,” so the love affair is real. Thankfully, camera brands see the same line in the sand: not one of them thinks generative AI belongs in a camera.

During CP+ last month, I had the opportunity to sit and chat with executives from every major camera brand and chat about a broad range of topics. Questions typically vary from company to company, but this year I asked every one of them the same thing: “Do you think generative AI should be added to a camera?”

I admit, this question was driven by a fear that one of them — and I wasn’t sure which — was going to try and cram AI into a camera body. Seeing what Caira did (even though it pulled back in response to feedback) had me nervous, and I feared that, given how much Western CEOs loved AI, someone in Japan was going to think this was a good idea. Luckily, at least for now, none of them thinks so.

Generative AI added multiple elements to this smartphone photo.

This might be the first time that Canon, Nikon, Sony, Fujifilm, OM System, Panasonic, and Sigma all agree on the same topic.

“Photography is not just about the result of the picture, but I think the joy of the process,” Sigma CEO Kazuto Yamaki tells me.

“So if we were just interested in the result, then AI-generated images can replace photos. But we enjoy using cameras, and we enjoy the process from RAW to JPEG. All this process makes the final result more precious. So I don’t believe photography can be replaced by AI-generated images, and I’m quite positive about the future of photography.”

“Nikon’s stance is straightforward: We will always back the human element of creativity as well as the photographers, filmmakers, and artists who bring it to life,” Fumiko Kawabata, Sr. Vice President of Marketing and Planning, at Nikon Inc., says.

“We need to collectively acknowledge the risks generative AI poses to both the concept and the spirit of originality, but also never forget that the passion for the process also brings us together.  Now is the time for us to be supporting our creator communities, and also exploring emerging tools such as content authenticity.”

The fireworks in the above photo were generated by a smartphone’s AI.

“Generative AI technology has pros and cons. We would like to develop some AI technology to deliver new value for the benefit of the customer. For example, noise reduction, image composition such as Live GND, or other functions like that can help expand the expression of the photographer,” Kazuhiro Togashi, SVP, Global Brand and Marketing Strategy at OM System, says.

“But, basically, we think the real experience of taking a photo is important. It’s a real photo.”

Adding elements to a photo, such as the fish in the above image, is a highly-touted feature of major smartphone brands. Camera manufacturers don’t think this is a good idea, though.

“We have to prevent the case where Generative AI starts to make fake images,” Go Tokura, Executive Vice President, Head of Imaging Group at Canon, says. “Photo authenticity is very important. Something that proves a photo was taken by our camera or disclosed that it was created by generative AI. ”

“I think what’s most important is to be able to distinguish clearly between whether it’s generative or not,” Yuji Igarashi, General Manager, Professional Imaging Group, Imaging Solutions Division at Fujifilm, says.

“I think at least as of now, we believe AI is something to assist taking pictures like autofocus, or detection, et cetera, rather than generating something in the camera. Any edits made with AI should probably be tracked better than what is being done now, so that when you have the file, you know what has happened to get to this at the end. So that’s our belief now.”

Fujifilm’s stance is pretty similar to those of its contemporaries. While cramming generative AI into cameras, similar to how some smartphone manufacturers have been so eager to implement, doesn’t sound right to these exectutives, they also recognize the benefits that the technology can have elsewhere. Sony, for example, has been using AI for years to improve its autofocus.

“When it comes to using AI, I see two main aspects. The first is camera usage: we actively use AI technologies such as human pose estimation and subject recognition. We fully embrace these technologies. At the same time, we want to protect and support creators and their rights,” Yasufumi Machitani, General Manager, Camera Business Department, Imaging Entertainment Business Unit at Sony, says.

“So, that’s why we offer Camera Authenticity Solution as a member of the C2PA steering committee, to support journalists and creators through more advanced verification of the authenticity of both still images and video content.”

Canon recognizes that Generative AI might also help its team’s efficiency.

“Generative AI may be helpful for enhancing the effectiveness and also the accuracy in the operation or operability of the shooting processes. There is one more possibility for internal costs as well. I mean, if AI becomes so useful for our development teams, for example, that could help us to reduce the cost, and then we can provide more value to the customers as a result.”

Panasonic says that it uses AI to improve subject detection accuracy or to allow a photographer or filmmaker to better express an image, but making something from nothing using AI is not the goal.

“Especially for video production, we feel that the subject needs to be real in that case. We focus on the professional live streaming systems, video logging, and daily use of the consumer camera. ‘Daily use’ does not mean generative AI,” Toshiyuki Tsumura, Executive Vice President and Director of Panasonic’s Imaging Business, says.

Every executive made it clear that while AI is hugely helpful for many aspects of photography, generative AI takes away from the artistic process that every company strongly supports. When I made it clear that the topic at hand wasn’t improving autofocus, adding in-camera computational features, or even improving resolution with upscaling technology, but rather what was being seen in popular smartphone brands that base their entire launches around creating fictional images and videos, the executives all nodded their heads in agreement: that isn’t photography.

“We have the same opinion about it,” OM System’s Togashi affirmed.

Image credits: Header photo licensed via Depositphotos.com.