How reporters outsmart the internet trolls
Illustration: Aïda Amer/Axios
Some of the top journalists covering misinformation today look less like reporters and more like cyber investigators, combining new technologies with old-school journalism principles to outsmart the trolls working to undermine them ahead of the 2020 election.
Why it matters: The troll playbook has shifted since the last election, when fringe internet actors sought to sow discord by spreading divisive messages. Today, their main focus is to discredit the news media, often by uncovering potentially harmful information about journalists or by tricking them into reporting false information.
Driving the news: New research from Data & Society by researchers Joan Donovan and Brian Friedberg lays out 4 ways that internet trolls aim to "source hack" reporters, or conceal the real sources of misinformation in order for it to go viral.
- Viral sloganeering: Repacking talking points so they go viral, such as "Jobs not Mobs."
- Leak forgery: Sharing forged media documents to prompt spectacle. The report authors note that leaks have been normalized in the wake of Edward Snowden and Wikileaks.
- Evidence collaging: Combining information from different, often not-credible sources, into an easily shareable social image that looks like fact-based evidence for hoaxes or conspiracies theories.
- Keyword squatting: Strategically making keywords go viral by using automation (bots) to spread misinformation from sock-puppet accounts.
What they're saying: Axios spoke with several of the top misinformation reporters today to break down the tactics they use to dodge these traps and identify the people behind them.
- Planting spies: There are so many remote internet communities where trolls congregate that reporters need to develop sources within those fringe communities to leak them information from within. "This is especially effective with platforms like Reddit that have very specialized communities," says Jane Lytvynenko, a reporter who covers misinformation for BuzzFeed News.
- Using reverse image search: Some of the trolls use the same avatars or images on multiple pages or social media accounts, says Ben Collins, a reporter on the "dystopia" beat for NBC News. Reverse image searches on Google make it easy to link those same pictures to one person.
- DM identity verification: "I recently tweeted at a QAnon character, and got a response. But in order to verify that the email response was actually from them, I asked them to send me a picture and to DM (direct message) me from their Twitter account so I knew it was actually them," says Will Sommer, a tech reporter for The Daily Beast and author of the Right Richter newsletter on conservative media.
- Studying language and punctuation: "We were looking in a suspicious Facebook page we thought might be Russian," says CNN technology and politics reporter Donie O'Sullivan. "So I downloaded all of the posts from the page and I started to notice that while the grammar was good, the posts consistently used the wrong type of apostrophe. So we called our Moscow bureau and asked to take a photo of their keyboards to see if they’re configured differently. Sure enough the strange apostrophe was on the Russian keyboard – after some more digging we proved the page was run out of St. Petersburg.”
- Following trolls early: "A lot of it is trying to see what might be coming my way, which means going to dark places of the internet and looking at campaigns as they start forming in the early stages, before they hit mainstream social media channels," says Lytvynenko. "You need to pay attention to key accounts during their early stages and then follow up and see how they act during breaking news situations in real-time."
- And seeing if they change their message: "You need to pay attention to key accounts during their early stages and then follow up and see how they act during breaking news situations in real-time." And seeing if they change their message: "You've got to keep an eye on accounts as they grow over years," O'Sullivan says. "We've found pages that looked to have an African American persona years ago tweeting about race and now are tweeting in Chinese about issues related to China.”
- Muting the trolls: "Muting trolls is important because it doesn't give them the satisfaction of seeing you block their account," says Lytvynenko.
- Refreshing pages: O'Sullivan recalls an instance in which he got off the phone with someone who denied running a fake news operation, and immediately hit refresh on the pages he thought that person ran, only to see that "a few minutes later I went back to a page and whoever was behind it temporary disabled it. It took months of reporting to prove it was him, but we knew it was likely him after that."
- Using fake phone numbers: "They're always different phone numbers. disposable numbers from throw-away services immune to this sort of thing," says Collins, referring to online trolls. As a result, he will give a specific number to people using those disposable numbers.
- Using encrypted messaging and emails app: Using encrypted messaging and email chat apps like Signal, Whatsapp, Viber and ProtonMail is crucial for almost all misinformation reporters we spoke to, and for many other reporters. In a recent interview for The New York Times, politics reporter Matthew Rosenberg says reporters need to go to extra lengths to ensure their phones can't be traced. "[Y]ou need to pay in cash, never register it with your own name or real email and never connect it to your home or work Wi-Fi network. I have two at the moment. One is a cheap smartphone, and the other is a flip."
- Measured responses: "You need to be extremely measured in your responses, because they want to call out malice on your behalf," says Sommer. "So it's best to respond to tips with things like 'Oh, interesting' or 'I would like to learn more.'"
The bottom line: "Our guard is always up," says Collins.