Doctored and Forged Screenshots: The New Weapon of Online Warfare

Popzazzle | Sunday, 21 March 2021 |

"Sparring Internet warriors have more recently taken to using doctored screenshots as a means to get their opponents' social profiles shut down... And in this twilight zone of low scrutiny, things can get very, very ugly."

Doctored Tweet
Tweet doctored by Bob Leggitt.

It's perhaps a stretch to say that doctoring a screenshot is the easiest thing in the world, but it takes less than a minute in Firefox or Chrome, with no additional software.

I'm deliberately not using the word “Photoshopped” in this post, because despite this being a common reference for edited screenshots, simple text-modification does not require anything as elaborate as a market-leading image editor. I'm using words like fake, forged and doctored, because most modified screenshots have not been anywhere near Photoshop. They've most likely been Chromed or Firefoxed. It's that easy.

And yet the result can convince a huge number of people that a completely fictitious statement or conversation was typed by the person to whom it's attributed. Let's face it, half the Internet will believe that Einstein's dating bio said “Genius. Swipe right” - if you put it on a suitably reverent background with a monochrome of his wizened face in the corner. And that's presenting someone's words secondhand. When you're presenting their words firsthand, the credibility rises further. And the stakes can be exceptionally high...

If, in the past week, you've seen mention of the Cardi B / Candace Owens debacle, in which a doctored screenshot was posted by the former, purporting to be a true representation of a Tweet from the latter, you'll know that a bit of imaginative type can blow up into international news. It's pretty well accepted that Cardi B did not forge the screenshot herself, and she would doubtless claim in any lawsuit pursued by Owens that she didn't know it was fake. But that won't necessarily absolve her of responsibility. And this was only the tip of a vast iceberg. The practice of faking screenshots has taken up a central position in online warfare.

On social media, there's an underground world in which screenshots are strategically doctored to achieve all sorts of ends. Everything from sploggers creating sensationalist celeb quotes as a means to drive traffic to ad spam dens, through fundraisers fabricating phoney outgoing expenses, to con men and predators forging astronomical bank and Cash App balances to bait targets. But the real groundswell in the usage of simple text modification has come in the field of character assassination and political smears.

Political smears are big business, and a large number of them have been supported by doctored screenshots. The power of political smear screenshots lies in the division of the public into two groups: support and opposition.

Broadly, everyone who opposes a political figure or party is going to accept that their adversary did something wrong, and shout about it, because it's in their interests to accept it and shout about it. That leads to rapid viral spread, which in turn leads to the “illusory truth effect”, in which even neutrals will come to believe something if they encounter it enough times.

But character assassination doesn't just affect major figures. It's also used in the small picture, against individual supporters of a party or ideological group. And whilst those individuals are frequently posting anonymously without any danger to their offline reputation, there's one point of attack where they're very vulnerable indeed...


It's almost impossible to debase small, anonymous individuals online, because they have no reputation, and virtually no expectations to live up to. And because they're part of a big gang, there's usually someone else around to argue in their favour.

For this reason, the standard ploy when an ideological argument becomes too difficult, is to attempt to get these small anons suspended from the platform on which the debate is taking place. Whilst this petty recourse to strategic tale-telling has been a thing since long before the likes of Twitter and Facebook emerged, social media has turned it into an art form. Some warriors in these online wars will even admit to reporting every tweet in an adversary's account. Just based on the time investment, that's pretty obsessive.

So perhaps almost inevitably, sparring Internet warriors have more recently taken to using doctored screenshots as a means to get their opponents' profiles shut down. Sounds unlikely at first, and you'd think it would be. But cross-platform posting, in conjunction with private messaging and very emotive, fake screenshots, can result, and indeed has resulted, in suspensions. It's a twilight zone of low scrutiny, where potentially, the only person seeing that fake screenshot is a moderator with no time to investigate. And in this twilight zone of low scrutiny, things can get very, very ugly.

Predatory pervert. Violent racist. Paedophile. The accusations are deliberately as dramatic as the forgers can make them.

You're obviously not going to get suspended from Twitter for something someone falsely claims you posted on Twitter, because moderators can access everything you've ever done on the site. But could you get suspended from an alternative social platform for something you're alleged to have said via Direct Message (DM) on Twitter? The answer is yes.

Some alternative platforms like to offer users of mainstream social the means to link their alt social accounts with a Twitter or Facebook identity. Can be useful. But once users of the alt site take up that offer, they can find the association being used against them by ideological opponents, trolls or stalkers. Once opposition warriors see that a confirmed link has been established on the alt platform, they report the user to the alt platform for something he or she allegedly did on Twitter. If they support this report with a completely uncheckable doctored screenshot - particularly a Twitter DM conversation - there's no way the alt platform can prove the exchange is fake.

This puts the independent platform in a difficult position. Whilst they can't prove the screenshot has not been doctored, it does make the report look a lot more convincing. I've seen examples of this, and it can appear very compelling.

If you run a social platform, and someone sends you screenshots of DMs in which a user of both your site and Twitter appears to be a threat to safety, what do you do? Without official intervention from law enforcement, Twitter won't confirm whether or not those DMs are real. And without any further info, the risk of ignoring a report like that is greater than the risk of acting upon it.

You can't just go on Snopes and fact-check a DM conversation. There are no independent witnesses. You can't reference it on Wayback Machine. You have no access to it. And text modification in-browser leaves no detectable signs of forgery (unless it's a Tweet and it exceeds the known character limit). So the decision on whether or not to act may come down to nothing more than user credibility. Is the accuser more credible than the accused?

What does seem likely is that an uncheckable, doctored screenshot could prompt much greater scrutiny of the accused user's behaviour on the independent platform. Even the slightest hint of dodgy behaviour could be enough to get them deleted, whereas without those dramatic, shocking screenshots, no one would have taken the trouble to look that hard.

We should categorise screenshot evidence alongside other propaganda tools like the “illusory truth effect”, as psychological trickery. Even when screenshots are not altered, they're often very selectively cropped to remove the bits that don't fit the poster's narrative. A whole sentence chopped off the bottom of a Tweet in some cases.

But we also have to be very careful not to hit the flipside, and automatically dismiss everything we suspect is doctored, just because it came from a source we don't see as reliable.

Ultimately, negative, text-based screenshots should at first glance be interpreted as an accusation - not evidence. If we can verify that the screenshot is genuine, then it's evidence. But failing that, and difficult as this may be, we have to assess the accusation without bias, as if the unverified screenshot did not exist.