Week14: Dangers of conspiracy theories in a global pandemic

Complaints by Topic

Through the Real411 platform, Media Monitoring Africa (MMA) has been tracking disinformation trends on digital platforms since the end of March 2020. Using the Real411 platform we have analysed disinformation trends which have largely focused on Covid-19.

The past week was officially the start of spring in South Africa. Sadly, the complaints submitted, while fresh, were anything but new and innovative. We have commented previously on the motivations of those who seek to spread disinformation and we have also looked at how disinformation isn’t necessarily home-grown but copied, usually from the US. This week is notable for two rehashed complaints, one about masks, (726) and the other about Covid-19 being a hoax (728).

Both complaints were reported as having been circulated via WhatsApp. WhatsApp presents a real challenge, both in terms of tracking down the original content, but also in curbing the spread of disinformation.

We know that 31% of all complaints received in the Real411 system have been found on WhatsApp. Unlike content on Facebook or Twitter or even websites, content spread on WhatsApp doesn’t have a URL (or Uniform Resource Locator). A URL is commonly referred to as a specific address where something can be found on the internet. When people ask you to click a link, or tell you to go to a specific website, what they are usually asking you to do is go to a specific address to see something. Facebook, Twitter and websites all have URLs that are easily visible. In the case of internet browsers, like the one you are currently reading this on now, it is going to be something like: https://www.dailymaverick.co.za/ with other bits after that. What that tells us is that we are looking at the site of Daily Maverick and then it tells our device where to find the specific article. 

So if you click on a new article you will see the first part of an address or URL will stay the same and the last part will change. URLs are useful not just because they tell us where to find things on the internet, they also give us a good idea as to what other content we might find and possibly who is behind it. Just by clicking on the URL for this piece, you can already see that you are on the Daily Maverick site. This is important in an age of mis and disinformation because it would be relatively easy to set up a site to look like a real or credible news site and then actually not be that site at all. 

One of the reasons banks, for example, constantly tell us when we sign in to be aware of fraud and phishing scams is that people will send us an email that looks just like our bank, and when you click the link it may even take you to a site that looks just like your bank — but if you check your address or search bar, the URL will not belong to the bank. If, for example, you bank with FNB, a dodgy email may well suggest you click the link and the URL may be: www.FNBB.co.za Unless you are paying close attention you might not see the additional B in the URL and think the address is credible.

Similarly, a few years ago, a few questionable people saw the gap and quickly set up sites that looked like credible media. There were sites that looked like Eyewitness News, SABC, News24, The Citizen and many others. Because they made the sites look similar, people often thought that they were actually visiting credible sites. Doing so meant users would be more likely to click stories that were not true and thus spread mis and disinformation. As we have seen with misinformation it is often created to make money. By creating a site that looks like News24, for example, and by getting lots of visitors to the site they could make a good deal of money from the advertising. 

Jean le Roux exposed some of these creators here. While he used a number of methods to identify the people responsible, one of the key pieces of information he looked at was the URL. URLs thus enable us to go back in time, see where something appeared and also track who might have put it up there. 

Let’s get back to WhatsApp: Unless you share a link to a URL in a WhatsApp message, the messages themselves have no visible URL for you to track or check. If you send an image or joke to your family and friends, there is no URL that we can use to tie that image back to it. The absence of an URL means not only is it more difficult to verify, but in the case of Real411 it makes it much harder to counter and if necessary try to remove. 

If, for example, a person shares child sexual abuse material from a website, the content can be traced, and the owners of the address can possibly be contacted and the content can be removed. You may have noticed a few riders in that last sentence, such as the owners or those who shared “may be traced”. Knowing the URL helps a lot, but those who create these sites know this too and use a number of measures to hide who they are, where they are and prevent people from tracking them.

On Real411, if content is found to be disinformation for example, or hate speech or incites violence, the secretariat can write to Facebook, Google or Twitter and ask them to please remove that post in order to prevent or limit harm. In the case of a website, if it is linked to a South African domain we can communicate the request to the ZA Domain Name Authority and they may take action. Without a URL, as is the case for the majority of complaints on WhatsApp, the content is almost impossible to remove. This reality means that not only is disinformation on WhatsApp extremely hard to trace and remove to help prevent further spread, it also means that content can just keep circulating, around the world.

The Real411 response is the following. If content like that in 728 is found to be disinformation, an infographic is created that outlines why it is disinformation, the complainant is then encouraged to circulate the infographic back on the same groups in which she may have received the original piece of content. In addition to this, Real411 circulates the same information to key stakeholders and partners who also help share the information to help curb the spread of the original. While it doesn’t remove the harm, it serves to mitigate it by at least ensuring that competing content is being shared. 

In the case of 728, as the complaint was assessed it emerged that one of the first times the image was used was in a Twitter handle in the US, and tweeted by that person in May 2020. What this tells us is that in the space of four months it has been circulated not just halfway across the world, but on a different platform. We have no idea how many times it may have been sent on WhatsApp and we cannot trace its journey. 

If we look at complaint 726, we see that it too comes from the US, its image being credited in a story in The Guardian from the end of July. So while more recent, we see it has taken content and added a call for a specific date as to when people will no longer wear masks.

Of course, as much as the technology may be a challenge, these things wouldn’t spread if people didn’t share them. Both 726 and 728 are clearly false. In the case of 728, the claim starts off calling Covid-19 a hoax, then escalates to anti-vaccination, microchips, a police state and surveillance. Despite being patently false, people still share this stuff. 

When we look at why, there may be a number of reasons, but the most common is that it resonates with people’s own internal biases and assumptions. These things are easy to see when they don’t fit our own biases, but a lot harder when they do. 

Our advice this week is: be sceptical not just of content from those you know you disagree with, but also always verify content from those you know you agree with. DM