Jane Lytvynenko is a senior reporter at BuzzFeed News, where she focuses on disinformation, cyber security and online investigations. She has uncovered social media manipulation campaigns, cryptocurrency scammers and financially motivated bad actors spreading disinformation. Her work also brings accessible fact-checking to wide audiences during times of crisis. Jane is from Kyiv, Ukraine, and currently resides in Toronto, Canada.

When news breaks, it can be hours or even days until reporters and officials are fully able to make sense of a situation. As early evidence and details begin to flow over social networks and other online platforms, bad actors can emerge to sow division or distrust, or make a quick buck off a worried news consumer’s attention. Those same well-meaning consumers and other sources can also unintentionally spread false or misleading information. The mix of heightened emotions and slow trickle of news in the early minutes and hours of an event makes it necessary for journalists to be equipped to effectively monitor, verify and — when necessary — debunk breaking news. A fake tweet, image, social media account or article takes just a few minutes to create, while real information struggles to keep up.

The key to monitoring and debunking during breaking news is to lay a strong foundation before it happens. This means having a solid grounding in basic verification skills, such as those outlined in the first Verification Handbook, understanding how to monitor social networks and platforms, and knowing how to respond if you or your colleagues become targeted by bad actors. Reporters should never put online safety on the back burner.

When news breaks, the first step is to identify key impacted communities. During the 2018 shooting at the high school in Parkland, Florida, reporters scoured the Snapchat map for videos of what was happening to the students trapped inside classrooms. In contrast, during Hurricane Irma in 2017, it was key to focus on Facebook, where those affected tried to find information. Understanding how each social network functions and how it intersects with a given event is essential.

This chapter will focus on tools a reporter can use for monitoring and debunking breaking news. Not every tool will be right for every situation, and understanding who has been affected can help you know which places to focus on most.

Three things to look for

As platforms and reporters work hard to fight disinformation, bad actors have evolved their tactics to avoid detection. Still, some consistent patterns of content and behavior emerge repeatedly.

1. Doctored or out-of-context imagery. The infamous image of a shark swimming on a flooded highway has been making rounds and continuing to trick people for years. (It was also the subject of a case study in the first Handbook.) Photos and videos from that have previously debunked are what fact-checkers and debunkers call zombie hoaxes and are important to watch for. Imagery spreads much faster across digital platforms than text, so focusing on them is often fruitful.

2. False victims or perpetrators. During the YouTube headquarters shooting, social networks were littered with false claims of suspects. During the U.S. midterm elections in 2018, false rumors about ballots being cast by illegal immigrants were spread by the U.S. president. False perpetrators show up during most big breaking news events.

3. Harassment and brigading. While not strictly disinformation, bad actors commonly try to harass people involved in a news event as a way of silencing them. It’s also a sign that a group of people is paying attention to an event and may try different tactics down the road. “Brigading” is when a group of people work together to create the impression of a groundswell of engagement or reaction, by doing things such as up- or down voting content or flooding a user with comments.

Best practices for archiving and publishing

Before looking for hoaxes, set up a folder for your documents and start a spreadsheet for what you find. Immediately take a screenshot of each hoax and relevant piece of content you discover, and archive the page. (The Archive.org web browser extension is a free, quick and effective tool for archiving content.) Be sure to record the original and archived URLs of the content in your spreadsheet. This enables you to come back to what you found and look for patterns after the dust settles.

To avoid helping spread pages associated with dis- or misinformation, be sure to link to the archived URL in any articles or social media posts instead of the original. It’s also a best practice to watermark your images with a clear label such as “False” or “Misleading” to ensure they are spread and indexed with the proper context. If you do write an article, focus your headline and copy on what’s true, rather than primarily saying what’s false. Studies have shown that repeating falsehoods can cause people to retain the incorrect information

Your role is to minimize the repetition of falsehoods as much as possible, and to steer people toward accurate information.

Identifying keywords and locations

As the event unfolds, come up with a list of locations and relevant keywords.

For location, take into account the city, state and country, and any relevant local terms such as the nickname for a city or affected neighborhood. During elections, you should also use the county or relevant electoral district name. This information is used to monitor geotagged posts and to search for mentions of the location. Also be sure to identify and begin monitoring the social accounts of any relevant local authorities, such as police and fire departments, politicians and local news outlets.

Next, identify key terms. This can include words like victim, suspect, shooter, shooting, flood, fire, the confirmed names of anyone involved and more general wording like “looking for” — think of the language people would use in the situation aside from key terms. If you find a credible account posting about being in the midst of the event you’re monitoring, note their username and read their full feed. Looking through their friends or followers list is also a helpful way to find others in the area who might have been affected.

Note that during stressful situations, people may misspell locations or names. For example, during the 2019 Kincade fire in California, some tweeted #kinkaidfire because of autocorrect issues. Include common misspellings in your searches and try to identify possible autocorrect mistakes by typing key terms on your device and watching what suggestions pop up.

This is also a good time to reach out to any sources you know in the relevant location or who are part of communities that might be targeted with harassment or disinformation, and ask what they’ve seen online. You can tell your audience that you’re on the lookout for disinformation and other problematic content related to the event. Coordinate with your newsroom’s social media team to help spread the word about your monitoring and to see if they have seen anything of note.

Key Image Tools

1. Image search

Reverse image search is an indispensable tool. It’s easy to search Google for an image by right-clicking on an image and selecting “Search Google for Image” in the Chrome web browser. But it’s always a good idea to search an image using different tools. If you install the InVID browser extension, you can right-click on an image and search it across different tools. This reverse image search comparison chart created by Domain Tools shows the relevant strengths and weaknesses of different reverse image products:

InVID

InVID is a free browser extension and the best platform for helping you analyze and verify videos. It allows for users to paste a URL into its engine, which will then extract thumbnails from the video. You can run reverse image searches on these thumbnails to see where else this video has appeared on the web.

2. TweetDeck search

The best way to search Twitter is by using TweetDeck, which allows you to create unique columns for searches and lists.

Finding and duplicating relevant lists is key for staying abreast of a situation. You can use Google to search for Twitter lists using a simple formula. Type site:twitter.com/*/lists into the search engine and then add a key word in quotes, for example “Alabama reporters.” The final search string is therefore:

site:twitter.com/*/lists “Alabama reporters”

This will bring up any lists that other Twitter users have created that include the phrase “Alabama reporters” in the title.

Once you’ve found a list that’s relevant for your needs, you need to duplicate it so you can add it to TweetDeck. Use this app: http://projects.noahliebman.net/listcopy/connect.php to duplicate as many as you like. It’s ideal to duplicate a list rather than to follow it because you can add or remove users as you like.

Along with finding and adding lists to TweetDeck columns, you want to create columns with specific search filters that enable you to quickly monitor for keywords as well as images and videos. To look for multiple keywords, wrap them in quotes and put “OR” between them, such as “Kincade” OR “Kinkade.” You can also exclude certain words if they produce irrelevant results. Most people no longer tag their tweets by location, so you can leave that field blank to cast a wider net.

If you want to narrow your results, set the “From” field to a day or two before the event took place, as this will make sure you don’t miss tweets because of possible time zone issues. If you’re still getting too many results, try filtering them by engagement to surface only the posts that others have liked or retweeted. You can also try breaking key terms into separate columns. For example, put locations in one column and other keywords in another. I usually break out a third column for possible names of suspects or victims and their misspellings.

Finally, if you’re seeing a very high volume of tweets, it’s a good idea to create a new column with your best keywords and to set the “Showing” option under the “Tweet content” filter to show only photos and videos. This will give you a feed that can help you spot viral or emerging visuals.

3. CrowdTangle

CrowdTangle is a web app and browser extension that’s free for newsrooms to use. (Contact the company if your newsroom is not set up with access.)

It’s a powerful tool that allows you to set up dashboards to monitor across Facebook, Instagram and Reddit. You can also search by keyword and set many filters, including time posted, language and engagement. CrowdTangle is especially useful for monitoring Facebook and checking where a URL may have been posted on social media. Once you have access, go to app.crowdtangle.com to get started and then click “Create New Dashboard.” Even if you don’t have access, the browser extension is free for anyone to use.

CrowdTangle: Searching for Facebook posts

Click on “Saved Searches” on the left sidebar and then “New Search.” You have two options with Facebook: search pages and search groups. I’d recommend doing both. Enter as many keywords as you like by separating them with commas. Then you can set how you see the posts, for example by most recent, most popular and overperforming, which is a measurement of posts receiving more engagement than is normal for a given page. I toggle among the three based on the situation to make sure I see viral content and new content.

You’re also able to sort posts by a specific time frame and type. CrowdTangle recently added the ability to search posts by the location of the page they were posted by. By clicking on “English” and then picking “Country,” you can select only posts that are coming from pages that have declared their location to be within the U.S., for example. You can also do the opposite and search for posts coming from pages based in Iran, Russia, Saudi Arabia, the Philippines or India, for example. Keep a special eye on image and video-based posts, which tend to spread further and be more engaging.

Once you’ve set up a search with relevant results be sure to save it so you can keep coming back to it.

CrowdTangle: Lists

Like TweetDeck, CrowdTangle allows you to build lists of pages and public groups of interest. By clicking “Lists” on the left sidebar and then “Create List,” you can monitor pages or groups that match keywords you have chosen or pages whose URL you have. CrowdTangle also has a number of prebuilt lists you can view by clicking the “Explore” tab. As with Twitter, building lists of pages and groups talking about the event you’re covering is a good way to monitor the information environment.

CrowdTangle: Link search

Another relevant CrowdTangle feature is link search. Go to https://apps.crowdtangle.com/search/ and paste in the URL or key terms of the content you’re interested in. CrowdTangle will show you the top public sharers of the link across Facebook, Instagram, Reddit and Twitter. (Note that the Twitter results are restricted to the previous seven days.) This will help you understand how the content is spreading, whether there are any groups or individuals you should be investigating further, and whether the content has spread far enough to warrant a debunk. There are no simple rules on when to debunk a falsehood, but some good questions to ask are: Has it spread outside of its initial network of sharers? Has it been shared by figures of authority? Has it generated significant engagement? (The free browser extension delivers the same data as the link search tool, and both are free for anyone to use without a full CrowdTangle account.)

4. Instagram.com

Instagram is a useful place to monitor for hashtags and geotagged posts. Look up relevant locations where users may have tagged photos, and remember that location tags can also include neighborhoods and landmarks. Once you found someone who appears to have been involved in a news event, click through to their account and make sure you watch their stories — they’re by far more popular than regular Instagram posts. Also look through the comments for other potential witnesses, and note any new hashtags that may have been used alongside their posts. If you want to archive someone’s Instagram story for your files, you can use a site like storysaver.net to download it.

5. SnapMap

Disinformation on Snapchat is uncommon, but its public map feature is useful to help verify or debunk hoaxes. To get started, go to map.snapchat.com and enter a location of interest. This will show you a heat map of where content is being posted — the brighter the location, the more Snaps are coming from there. To save a useful Snap, click on three dots in the top right and select “Share.” You’ll be able to copy the URL of the Snap to look at later. (Be sure to screenshot it as well.)

Putting it all together

It’s essential to practice using each tool before news breaks to avoid scrambling in the moment. Disinformation is meant to play on emotions and capitalize on gaps in news coverage. Keep that in mind as you search the web. You will also often come across accurate information that could help your colleagues. Write down everything that you know is true so you can recognize false things faster, and don’t be afraid to ask any reporters your outlet has on the ground for help.

After the dust settles, it’s helpful to look back at your saved images and posts. While in the moment you want to highlight individual falsehoods by way of public service journalism, in the aftermath you should take stock of any themes or patterns that can be seen. Were people targeted for their race or gender? Did hoaxes that originated on small, anonymous accounts become mainstream? Did any social media companies perform especially well or especially poorly? A wrap-up story can help your readers fully grasp the purpose and methods of the disinformation’s spread. It will also serve as a research tool for you and your newsroom, showing you what might be useful to focus on the next time news breaks.