We all know what fake news is. A phrase that ballooned to immense proportions during Trump’s presidency, the covid pandemic and the war on Ukraine. Social media, influencers and the anonymity of the internet made the dispersion of rumours and information seemingly impossible to control or regulate, or so it may seem. However, one must ask oneself how and through what means fake news is dispersed. Can we find patterns to understand the causes, drivers, and impacts that fake news truly has on our society in these times of crisis and turmoil?
TIM’s Data Room conducted a digital analysis for the Italian digital media observatory, bringing us some interesting findings. Their published document entails a digital analysis of the ten most diffused items of fake news in the Italian internet sphere in the month of May. By using listening tools of these platforms, cross-referencing keywords, and employing image recognition, a large information base was established. One important omission in the platforms observed was Facebook. This is due to the lack of tools to access the platform’s data. This may have skewered the results somewhat, as Facebook is the most populous platform and has a lot of media traffic.
The ten most diffused Fake News Stories are the following. These were broken down thematically into recurring Fake News topics.
War
1. Ukrainian band Kalush Orchestra gives the Nazi salute at the Eurovision song contest
2. Lavrov says: Hitler had Jewish origins
3. Finland has sent troops and tanks to the border with Russia
4. Poland wants to annex part of western Ukraine
5. Photo with overturned cars but undamaged windows shows that it is a staged
by Ukrainians to misinform about the Russian military invasion
6. The Bild publishes an article about Ukrainian refugees setting fire to the home that houses them
Health
7. Bill Gates predicted the monkeypox pandemic
8. Monkeypox cases are caused by Covid vaccines
9. Pfizer CEO wants to reduce the world population by 50% by 2023
LGBTQ+
10. The massacre in Uvalde, Texas, was the work of a transgender woman
The data behind these fake news stories and conversations can be broken down into many metrics. We can look at volume, spread, demographic origin and even the relationship between these different fake news stories. Let us start with origins. 7% of these stories and conversations originated from blogs and forums. 12% came from online news outlets. Finally, 80% originated from Twitter. Twitter was the focal platform on which most of this data was based.
A small minority of only 12 000 people were individual fake news authors. Roughly double where fake news stories emerged from conversations. These appeared 118 Million times on people’s Twitter timelines. Of this enormous number, 228 000 were engagements (comments, shares, etc..). When seen isolated and out of context, this data doesn’t mean much to most of us. We will break it down for you in the following steps.
These numbers mean that a small number of people have created fake news. The story then gets accelerated and reaches a high initial peak of viewership and circulation. Although the fake news reaches a lot of people’s timelines, it is shared and diffused by a small portion of these viewers. This news spread is thus accountable to a small number of people rather than everyone the news reaches. Once Fake news is disproved, it loses its circulation strength.
An example is the story about the Ukrainians doing a Nazi salute at Eurovision. Many assume this was Russian counter-propaganda, and it quickly lost its momentum. Some news, however, bucks this trend. The fake news about Poland wanting to annex part of Ukraine is still experiencing peaks in circulation.
Another critical pattern found is that an average of 5.1 days pass between disseminating and denying a fake news story. This may not sound like much, but it is enough time to reach hundreds of millions of users. Of course, people with high centrality in a social network (many followers, in many pools) can be catalysts for circulating fake news.
There is some positive news, too, however. A majority (56%) of people who commented or reacted to fake news were actually denying the news. Once more, this is highly dependent on who diffuses it and what places it reaches. Russian foreign minister Sergey Lavrov has 94 per cent of his followers who defend his claims. Elsewhere, fake news would encounter far more resistance.
As aforementioned, demographics are key elements in understanding how and why fake news moves across our networks. A very paradoxical finding was that men are more engaged in both disproving and spreading fake news. 56 % of diffusers of fake news are male, and 67 % of deniers are male. The age group most active in diffusing and denying fake news is the 25 to 34-year-old demographic.
We haven’t mentioned influencers yet, and the power people have inside the network. By conducting a detailed social network analysis (NSA), the research found a structured relationship between the different and most popular fake news stories. Individuals are represented as nodes, and we visualise the connections of social media platforms like Twitter by connecting these nodes. You start seeing communities by seeing all the relationships between people on social media. These are places where nodes (users) have a larger amount of connections between each other. The more connections, the stronger the community (see visual). People whose node is very central in the community could be considered influencers as they hold a key position in how information travels.
The study showed that a very strong community existed around health-related issues and the topic of vaccines in particular. This results in a very fast diffusion of information. The topic of war has a much larger and weaker community with a lower density of connections between people. This means it can be disproved faster whilst more fake news circulates across a broader spectrum. If we take a step back, however, we see that all of the fake news stories selected are heavily connected regardless of the topic. People who are receptive to getting fake news about one topic are more likely to also receive other fake news items. Fake news communities do form around certain topics, but there exists a significant overlap between fake vaccination news, war fake news, refugee fake news and many others.
This research gives us numbers to try and understand the behavioural patterns of fake news on the internet. How they spread, how they die, and who is responsible along the way. The nature of social media means that as a bystander or opponent, you can still help popularise fake news even without sharing it. One positive conclusion is that once fake news is disproven, it rapidly declines in popularity. This means it can be combatted effectively. However, it also shows that people who are receptive to fake news in one area are likely to be affected by other fake news too. We must shorten the time it takes to deny and shut down fake news. After all, it is a minority that creates this news, but it is the structure of social media such as Twitter which provide it with fertile ground for circulation. If you react in favour or against the fake news item, you are still helping it diffuse by dispersing it throughout your network. If we must draw some conclusions, we must be careful and responsible in our online space. Fake news and the way it spreads across the internet to infect many people’s minds is no mystery. Research like this helps us think about solutions to fake news and its divisive and destructive wake.
