The Disinformation Problem
Updated: Nov 24, 2020
In 2016, with the help of the DHS and FBI, the Senate Committee for Intelligence released an exhaustive report on a Russian campaign to influence the Presidential elections through disinformation on social media. Since then, numerous reputable news organizations have agreed with our intelligence services that foreign disinformation, be it Russian, Chinese or Iranian, has circulated in our social media networks to the detriment of open and authentic conversation.
The question of state-run disinformation campaigns can feel distant and abstract to the average American. Unlike the “fake news” label popularized by the current President, “disinformation”, and its close cousin “misinformation”, are not intuitively understood.1 If you happen to think that CNN peddles fake news, an easy solution would be to simply stop watching the network. If you read a report about a complex foreign disinformation campaign sweeping Facebook and Twitter, your solution can hardly be to abandon social media. Here at dKomplex, we’d like to bring you up-to-date, easily digestible reports on the disinformation situation in your area, starting with Monterey County.
With both a nation-wide movement against racial injustice and a pandemic in full swing, malicious actors both foreign and domestic have plenty of ammunition with which to disrupt our social media conversations. The dKomplex team is concerned about how this disinformation can seep into and destabilize the social media landscape at the local level. For an example of how a single piece of content can cause real harm, observe the impact a fake protest flyer circulated on Facebook had on Watsonville on the June 6th weekend. Although Watsonville PD quickly realized it was a hoax, the damage had been done: small businesses had boarded up their properties in anticipation of the “barbaric protests”.
In an effort to better understand which user networks in the County are most vulnerable to this kind of content, we took a look at open-source data from the County’s Twitter network for the month of March. If you are interested in reading the full report, you can access it here. Here are some of our key findings.
When we look at a Twitter network map of the County, most users exist in the “middle” of the web, what we call the main component, while clusters of users reside in the peripheral component. The good news is that the main component is mainly populated by mainstream news sources, public officials and academics. Unsurprisingly, these public figures represent the largest (most interconnected) nodes in the map. Users in this component are exposed to authentic sources of information. The troubling news is that they have little connection to the peripheral clusters, who form echo chambers that can amplify misleading content. It is in these clusters where we find the most interconnected and politically active users. Influential users are not in themselves a source of concern, but when they have influence comparable to KSBW news within their networks, they can easily spread misinformation.
Though we have not found any calls to violent action or similarly extreme content, several examples of both Covid-19 and protest disinformation jumped out at us in our research. The following tweet is from one of two prolific users who we’ve identified as retweeting Covid-19 misinformation.
“As I've said numerous times don't #donttrustmedia @CNN @PressSec So why would you trust the numbers when it comes to #COVID19 ... Say goodbye to #USeconomy & hello extreme #USpoverty smarten up people! #misinformed #LiesLiesLie”
Misinformation that we found wasn’t restricted to the pandemic. Some local users were spreading doubts about George Floyd’s autopsy results:
“We now know you can’t trust official medical examiners”
These tweets were shared by an out-of-state user with over 300,000 users, thus suggesting that rumors born in relatively small local networks can spill into the national conversation. The reverse is also true – once content enters a local echo chamber from outside, it is unlikely to be challenged.
The term “Russian bots” has become popular since the last election cycle, but automated accounts are not limited to foreign actors. Bots are a significant problem for Twitter – some studies demonstrate that 15% of Twitter accounts exhibit signs of automation. Thankfully, we found a much smaller proportion of suspicious accounts on the Monterey Twitter network. Some of those accounts simply retweet other news outlets at a consistent rate, while adding no original content. Sometimes these “newsfeed” accounts share medically accurate content from public health sources, but other times they pull data from dubious news websites. Of greater interest are those conspiratorial accounts that are being retweeted by other suspected bots. Though the original account may not necessarily be a bot, their content can be amplified far beyond their small following by automated retweeting throughout the network.
The Bottom Line
At a time when false rumors can lead to serious consequences, it is more important than ever to pay attention to the news we consume. Even when it became obvious that there were no antifa buses on route to the small town of Bethel, Ohio, some 200 counter-protestors continued to violently disrupt what was originally a peaceful, 50-person demonstration.
At dKomplex, our goal is to uncover the truth to protect democracy. We think that the people of Monterey County deserve a clear and easy-to-understand picture of our social media landscape so that we can protect ourselves from misleading content. Stay tuned to the dKomplex Research blog for further updates on the health of our digital community and check out our full reports for the complete picture.