When some people think of a fact-checker, they only see a journalist who looks into a claim, checks it and confirms whether it is true or not. Nevertheless, fighting disinformation effectively now a days takes much more than that: quickly detecting possible hoaxes before they go viral and affect more people; explaining that not everything is black or white, helping people to be well-prepared against falsehoods through Media Literacy or regulation; creating tools to make it harder for others to deceive us, or fighting together against manipulation… In Maldita.es we make all these things to prevent people from being fooled: this is what we call pre-bunking.
Early detection of disinformation contents and narratives
In Maldita.es we are working on different projects to obtain early warnings of potential disinformation. We use algorithmic tools to learn from the language employed by those disseminating disinformation so we can automate processes and create systems that allow us to be faster when reacting against disinformation.
Automation and chatbots
Through our chatbot service in WhatsApp (+34 644 22 93 19) we are more efficient at various aspects of pre-bunking. On one hand, people send content to us to ask whether it is false or what we know about it. The chatbot automatically replies with articles already produced by Maldita and, if that is what the user was looking for, asks them to forward the link to those they received it from, ensuring that it spreads over private networks and that others are protected as well. In addition to that, those queries provide us with real-time information about what is going viral online and, above all, they help us discover new disinformation contents and narratives that can be part of organized (and even foreign) disinformation campaigns.
In Maldita.es, the community plays a key role in detecting, debunking and sharing debunks in a quick and efficient manner. As we mentioned, users can send us content they are doubtful of not only via WhatsApp but also through other social media platforms such as Twitter, Telegram, Facebook or Instagram.
Our community is also part of our pre-bunking efforts through our ‘Superpoderes’ initiative. Experts in different fields voluntarily participate in this program, allowing us to understand faster the pieces of content we debunk and include them as reliable sources. Also, when they share articles and debunks they have collaborated in, we are able to raise awareness and prevent other professionals from engaging on disinformation narratives that are being spread in their areas of expertise.
Education is an essential ally in the battle against disinformation. In Maldita Educa, we develop materials, videos and training tailored to meet each person's needs that explain how disinformation works, what motivations are behind it and how to fight against it. Not only is it crucial to discuss misinformation with people, but it is also beneficial to teach them how to look for and consume reliable information and promote critical thinking.
All of these pre-bunking methods are used by Maldita.es. By combining them, we aim to increase awareness and assure people have access to enough resources to fight against disinformation when they stumble upon it. This is how we vaccinate society against the disinformation virus.
On the other side, there are those (typically some big online platforms) that are always looking for a magical remedy, a single and final solution against disinformation, as cheap and automatic as possible. A “silver bullet” that never arrives, but enables them to avoid taking actions that would greatly help against disinformation, using the search for an ultimate solution that does not exist as justification. A solution that YouTube now refers to as “pre-bunking”, but that it is significantly more constrained than what pre-bunnking actually is.
Inoculation theory as a part of pre-bunking
The part of pre-bunking that platforms like YouTube thoroughly promote is a more limited approach based on the inoculation theory. In the disinformation field, it would mean that before encountering a hoax, people are given arguments to refute it, either specific information about why it is false or (as some platforms prefer) generic warnings about the techniques that disinformers often use.
For example, it can be carried out through online games such as Go Viral!, focused on disinformation about Covid-19 and which we already talked about in this article on video games. Maldita.es has also applied the inoculation theory to pre-bunking in projects such as the escape room “The Hoax Factory” on migration misinformation. There are several studies in which this strategy has had positive results after being applied to topics like climate change disinformation and to different formats such as short videos.
That being said, in Maldita we believe that fighting disinformation using a pre-bunking strategy based only on these inoculation actions is doomed to fail. A wide variety of tools are needed, a much broader concept of everything that constitutes pre-bunking: early detection, education, community... in addition to debunking, that various studies show to be more effective in the long term.
Why are platforms so enthusiastic about this approach? The example of Google in YouTube
If debunks have better results than inoculation initiatives there are two main reasons why some big-tech companies are more interested in the second option: cost and scalability.
Monitoring specific disinformation that circulates online and investigating it with ethical and methodological guarantees requires a group of professionals who detect, investigate and prepare specific information for each piece of disinformation on a daily-basis.
If we take as an example an initiative by Google in which they created brief videos that describe methods frequently used to spread misinformation and placed them before the video containing misinformation, we can see that inoculation is cheaper in that case: very few videos are produced and they are not content-specific, but generic. In this way, they also reach a very wide audience, since they deployed the animations as ads on the platform, and could be placed next to many different types of videos since the YouTube project avoids getting into specific narratives.
On the other hand, Google has used a different strategy in Eastern European countries in order to dismantle “emergent narratives” about Ukrainian refugees (e.g. increase of living-cost is to be blamed on ukrainian refugees) that will be soon extended to Germany given the “promising results” that Google itself reports. The detection of these emerging narratives was a task conducted by fact-checking organizations such as Demagog (Poland), among others. The inoculation therefore depended on the daily work of detection and verification carried out by the fact-checking organizations.
While YouTube sticks to this narrow and incomplete approach to pre-bunking, which it merely applies to a few experimental applications with a very limited reach, the problem of disinformation and radicalization on the platform remains very serious. In January 2021, more than 80 fact-checking organizations from all around the world sent a letter to the company demanding action to stop the platform to allow it being “used to manipulate and exploit”, but there have been no significant changes.
In the fight against disinformation, inoculation needs to be accompanied by other efforts, focusing only on this approach because it is cheap and scalable can be dangerous. There must be safeguards in place to tend to those who have already been exposed to disinformation without having been inoculated with pre-bunking, both because they did not want to or because they did not have the opportunity, or for those in which the action has not worked as expected.
It has been demonstrated that the inoculation is more effective if it is developed in a more specific way and tailored to the different groups of users. This requires extensive collaboration between fact-checking academics, platforms and organizations to craft the most appropriate messages. It must also be taken into account that the results of pre-bunking in different demographic and cultural groups can vary, for instance, researchers attributed null results in an inoculation intervention carried out in India to “cultural and experimental design factors”, as the vast majority of studies have been done in Western countries. Also, there is much to investigate about the duration of the effect and the need for "booster doses" since the effects in time are limited, as Van der Linden explains.
Even though this inoculation theory was first formulated in the 60s, the reality is that there is not enough evidence on the results of these initiatives when in the real world, integrated into platforms such as YouTube. In addition, the investigations that are carried out will necessarily have to be in collaboration with the platforms since they are the ones who have access to the complete data.