Seeing isn’t always believing: How to spot fake, edited images online

The Right Now MN PAC has taken down this post.
Here are some ways to spot images that are forged or manipulated.
Jiwon Choi | MPR News

This is part of “Can You Believe It?” a series of stories and resources focused on giving you the tools to fight disinformation heading into Election 2020. You can find more resources here. What have you been seeing? Share here.

Photoshop and similar image editing software have become so sophisticated, it’s difficult to tell what’s real and what isn’t when it comes to visuals on the internet.

“Historically, [it] has been said that evidence can be difficult to believe unless it’s visual. So, if you have a picture, historically, you can’t refute a picture,” said Mike Johnson, director of graduate studies for the University of Minnesota’s security technologies master’s program. “We're kind of sensitized to that and we kind of jump to the conclusion that if it's an image, it must be real.”

Of course, that’s not true. And it hasn’t been for a long time — photo manipulation is as old as photos themselves.

Create a More Connected Minnesota

MPR News is your trusted resource for the news you need. With your support, MPR News brings accessible, courageous journalism and authentic conversation to everyone - free of paywalls and barriers. Your gift makes a difference.

Technology has improved by leaps and bounds in recent years, and high-quality image manipulation is difficult to detect. MPR News contacted two experts — Johnson and Mona Kasra, assistant professor of digital media design at the University of Virginia — for advice on how to spot fake images.

Using your eyes

Johnson and Kasra recommend examining pictures closely for clues. In doctored pictures, reflective surfaces like mirrors don’t change, some objects may appear smaller or larger than they would in real life, and shadows will be inconsistent. Pixelation or other digital artifacts are also a giveaway.

Using your brain

Disinformation relies on our emotion and confirmation bias. We’re more likely to believe something is true if it aligns with our own beliefs, and vice versa.

Kasra studied fake images for several years and said people’s opinions are a major factor in their ability to discern real from fake images — and a big reason they share them.

“You can get people to take action from an emotional perspective if you give them an image that's really upsetting,” Johnson said.

Of course, good designers and photographers make images that may elicit an emotional response. That’s part of their job. But if an image elicits an extreme response, evaluate it more deeply: Question its maker’s intent, its source and evaluate it for the rest of the factors in this article.

Johnson also says context is important. If it’s not plausible, it’s probably not real.

“Do you really believe that the president of the United States had lunch with an alien? That's something you can automatically just move past because there's not even worth looking into,” Johnson said.

Using technology

Technology to spot manipulated images is “always one step behind” the technology used to create them, Kasra said.

Nevertheless, Johnson says to check the metadata — a cache of information stored in an image’s file — for clues.

“Details about the file, where the picture was taken, when it was last saved, what the copyright [is], what the format of the file is,” he said.

Many computers will allow you to examine metadata without any software, but there are free programs available, if needed.

There’s one major disclaimer with metadata, though: It can be manipulated, too.

Another approach is a reverse image search that allows you to compare potentially manipulated photographs against real ones. These tools also show you what person or organization created the photo, and where else it has appeared online.

Finally, new tools on the internet allow you to analyze images.

FotoForensics detects anomalies in images, suggesting areas that may have been manipulated. Forensically offers error level analysis plus clone detection and several other analyses.

But no tool can accurately and definitely say an image is real or fake, so use only with that in mind.