When a friend or family member shares an image on their timeline, do we ever wonder where it came from? Or even, if it’s real? Some photographs are clearly too good to be true, but a sucker retweets every minute, and there’s a lot of money to be made from our collective gullibility.
Who out there can help us find fact from fiction in online photos?
Paulo Ordoveza hunts down faked and stolen photos on Twitter under the name @PicPedant. He calls out uncredited photos like he sees them — a practice which naturally peaked Pixsy’s interest.
Self-described as a “Punctilious internet killjoy at the forefront of the New Debunkonomy”, Paulo is obsessed with “attribution and Photoshop”. With 23.1K followers, he makes a rather entertaining sport out of back-linking stolen images and alerting their rights holders.
Here’s his game-plan: when an anonymous image starts racking up likes and retweets, he finds and tags the original creator.
— PicPedant (@PicPedant) January 24, 2017
While thankful, the photographer’s reaction is usually one of jaded resignation.
“Any photographer who’s seen a photo go viral on Reddit know that a flood of bot copies come next,” he tells Pixsy. The ones illegally posting the photos, on the other hand, rarely take notice of @PicPedant’s callouts.
“They usually put “parody” or “fair use” in their bios, which they seem to think are magic words against any infringement claims,” he says. Like many infringers we encounter, there’s a purveying ignorance towards posting photos uncredited. “One went so far as to block the original photographer when he complained.”
And who posts these photos? Typical of the digital age, the culprits are rarely human. Your average picture spammer is merely a marketing program, designed to algorithmically recycle old photos for likes and shares. The content is scraped from Reddit or Imgur, then shared to thousands via an RSS schedule.
Spam accounts require little to no maintenance once they get going, so there’s no one behind the desk to confront. It’s unfortunately why these pages are often unsuitable for Pixsy’s resolution process.
— PicPedant (@PicPedant) December 31, 2016
But even if the photo thief is just an unwitting AI, @PicPedant and other debunk/attribution accounts can still educate the unsuspecting users who help share these images en masse. That is if they aren’t fake too — companies usually jumpstart their Spammer accounts by bulk-purchasing a few thousand followers, to pass themselves off as popular.
From Paulo’s experience, even those who find their following organically do so by mass-plagiarizing memes and captions off Reddit. “Copy-pasted pop culture wit seems to outweigh accuracy for viral lift,” he remarks. When you can’t afford billboard space on the social media highway, plagiarism is a good plan-B.
When he’s not finding a photo’s rightful owner, @PicPendant devotes time to debunking fake and hoaxed images on social media. He’s not alone in this endeavour either.
As his handle would suggest, Janne Ahlberg a.k.a @HoaxEye shines a revealing light on viral images that either don’t show what they claim, or are entirely fabricated. A cleverly manipulated photo can fool anyone, but while the technology churning out fake images is sophisticated, the photos are anything but.
“I’m not sure if all people sharing fake pictures or stories actually believe in them,” says Janne. “Some might retweet or like a fake picture because it “looks nice”. Perhaps some people don’t actually care if a picture or story is real.”
Indeed, many of them aren’t. Ever marvelled at a crystal-clear view of the galaxy? How about the frozen canals of Venice, which haven’t actually frozen over since 2012? These images are usually CGI-renders, snatched from the defenceless portfolio of an inspecting graphic artist.
— HoaxEye (@hoaxeye) March 17, 2017
— PicPedant (@PicPedant) February 10, 2017
Of course, you don’t need to be a professional to pull the rug under Twitter users. Photoshopping one element into the photo, such as a celebrity, or inflammatory slogan, will also do the trick.
But why even bother with that? Slide a subtle lie into your 140 characters, and you can send military platoons to countries they’ve never set foot on, or transform movie stills into real historical photographs. Context, as they say, is everything.
While there is meaning in it, it's just all wrong. Time and place altered to suit a fabrication. And of 120k followers, many will accept it pic.twitter.com/Sjc7aYIMwn
— Andy Arthur (@cocteautriplets) February 18, 2017
— HoaxEye (@hoaxeye) February 22, 2017
Once you make a benign image breathtaking, it’s easy to see how fragile human credulity really is. “Basic fact checking is not very difficult,” says Janne. “You could easily catch over 90% of fake pictures by using reverse image search tools or fact checking sites.”
Some images are so farfetched, it’s hard to believe anyone could believe them. Asked for instances where even he was surprised, @Hoaxeye listed the following:
- Despite obvious Photoshopping, this downward “aerial” shot was too sophisticated for Nikon Singapore, who awarded a prize for its perfect timing.
- A shot of Eddie Murphy eating stake and chips off a naked model. The photo is real, but the Eddie Murphy is not.
Or, how about this shot?
A basic grasp of American history should tell you that Barack Obama was still a child when Martin Luther King Jr. and Malcom X were assassinated. Nevertheless, Snopes had to explain why this viral “photograph” of all three figures, sitting together in the prime of adulthood, was chronologically impossible.
“It’s not about realism so much as it is about the emotions a scene creates in them,” said @PicPedant. “Emotions that pic spammers then monetize.”
— PicPedant (@PicPedant) March 9, 2017
In 2017, debunking fake images has become more vital than ever. Now it’s not just landscape, nature or historical shots where truth is twisted for profit. Websites with extreme political views and vested interests are mass-producing doctored images on a daily basis, and relying on visceral outrage to fuel their monetization models. It’s the cornerstone of the emerging “Fake News” industry.
“I used to wonder how cultures of conspiracy developed in otherwise well-educated nations,” comments @PicPedant. “2016 was a learning experience in seeing how wealthy authoritarians weaponized disinformation networks to achieve certain political ends.”
Remember the infamous “Pizzagate” conspiracy theory in 2016? In this case, a string of fake news articles culminated in a gun attack in a restaurant. Could a fiction created in a hoax photo lead someone into danger?
@Hoaxeye singled out an image of a U.S. SWAT team, passed off as “foreign mercenaries” sent to intervene in Ukraine. Likewise, @PicPedant recalls where dramatic images of Brazil’s 2013 mass riots were passed off as new images of Venezuela’s 2014 government protests. This kind of misinformation could very well have incited further violence on the ground.
“Fake news has been always around – same goes for doctored photos, conspiracy theories, propaganda etc.,” says @HoaxEye. “On the bright side, social media giants like Google and Facebook are turning to fact checkers to fight fake news.”
Until social media sites develop a robust solution to fake images, it’s down to the debunkers. Last year, Snopes partnered with Facebook to take on the “Fake News” phenomenon, and there are many alternative organisations, including TruthOrFiction.com and Vmyths.
Twitter and Tumblr are full of debunking accounts, but as a movement, it currently lacks coordination. Paulo has toyed with the idea of creating a @PicPedant Subreddit or Slack channel, but as it’s not his day job, he says he lacks the time to deploy a professional service.
So how can you tell if a photo online is real?
Always start at the source. Especially if it’s a low-res image that’s obviously been cropped.
Always distrust social media accounts which share photos alongside spammy affiliate marketing links. Finally, unfollow any page with “Porn” in the title (e.g. “Earthporn”, “Historyporn”), unless followers of the group are the ones providing content.
If you want to get into the debunking/attribution game yourself, here’s a few starter tips:
- Use reverse image search tools, such as Pixsy, to find out where the image has come from, and how it’s been changed.
- Try running a description of the image through various stock photography services. This is often the source when a photo claims to show something it actually doesn’t.
- Do a phrase search on Twitter, Reddit etc. to see when jokes or captions have been recycled.