Translation Results Requested Through Conyac Made Public

This Startup Can Tell You if a Photo Is Fake

In the age of Photoshop, it’s almost impossible to say with absolute certainty whether any given image is real, but one former Photoshop employee hopes to change that.

Fourandsix Technologies, a startup founded by a former Adobe Photoshop executive and a digital forensics expert, unveiled its first piece of software this week, which promises to help law enforcement determine whether a photo is authentic or not. The software, called FourMatch, analyzes the meta data in picture files to quickly determine if a photo has been modified. FourMatch relies on a comprehensive database of more than 70,000 “signatures” that are left on a file from each piece of hardware and software that goes into creating it. As the company explains in a description of the product, “Once an image has been edited and resaved from a software product, this signature is changed to match the software rather than the original capture device.” FourMatch is primarily intended for police and lawyers who need to determine whether a photograph has been tampered with in any way between the time it was first captured and submitted as evidence. However, the software won’t tell you exactly how the image has been altered — if it has been at all — because it only analyzes the file data rather than the image itself. For this reason, it can just tell you if a file has been touched by another application. “This first product we’ve put out is not a magic bullet that will tell you everything you need to know about an image,” Kevin Connor, the company’s president and co-founder who worked at Adobe for 15 years, told Mashable. “This is sort of a first step and there are certain scenarios when it will be very valuable, particularly in the law enforcement space.” Indeed, the software currently retails for $890 so it’s clearly not intended for the average consumer. In the future, though, the startup plans to release other tools to determine the authenticity of pictures that should have broader use. In particular, Connor sees a growing need for technology that can detect photo fraud in medical research, help media companies assess whether their photographers have been too liberal editing their photos and sift through pictures going viral on social media to figure out whether they’re real. Likewise, he thinks tools like this could help banks verify the authenticity of pictures of checks and other payments taken by customers. “People are using images more and more to communicate and facilitate transactions,” Connor said. “There will be more situations when you might want to be able to verify that it’s true.”