It's no secret: for a number of years now, driven by the explosion of social networks and changing habits, false information has been polluting the information ecosystem. One could even speak more broadly of information manipulation because alongside false information, there are manipulated facts, partially false information, misleading and biased information. Although they have always existed, false information has truly gained significant ground. In the past, those wishing to spread false information could only rely on confidential publications, photocopies... Today, with the ubiquity of social networks, they can quickly reach a fairly large audience.
However, estimating the proportion of false information is extremely difficult given the extraordinary power of platforms. Facebook, Instagram, WhatsApp, Twitter, TikTok, Snapchat, private messaging... Several billion people use social networks worldwide. False information is everywhere and follows the pace of current events. The hotter the news, the more prevalent they become.
Why did Agence France-Presse decide to venture into fact-checking?
Our mission, always, has been to disseminate verified information. One might wonder why AFP is doing fact-checking verifying facts since it's the very essence and promise of AFP to its clients. There's an explanation for that. Of course, AFP verified facts before, but until recently, it didn't address rumors and false information.
2016 was a pivotal year for AFP and all media. In two major events the Brexit in the UK and the election of Donald Trump in the United States false information played a role. This realization of the significant role false information played in public discourse led AFP to take action. Today, when you're a modern news agency, when falsehoods are practically consumed as much or even more than truth, you cannot just stand by. We decided, like others of course we're not necessarily pioneers in this that we needed to focus on working with false information and explaining why they were false, partially false, or distorted.
How does AFP implement fact-checking?
Often, an event leads to launching a project. For AFP, this event was the 2017 French presidential elections. We started digital investigation with just one journalist. Today, we're 130, spread across around thirty bureaus worldwide. AFP fact-checking journalists are present on all continents: in the US, Argentina, South Africa, Thailand, Australia, Germany... We work in 24 languages, from obvious ones like French, English, Arabic, to less obvious ones not because they're more confidential, sometimes they're spoken more than French, but less obvious like Thai, Burmese, Slovak, Finnish, Catalan... Today, AFP has the largest global network of fact-checkers among media outlets.
You mention 'digital investigation' rather than just fact-checking. What's the difference?
Beyond fact-checking, within AFP, we conduct digital investigations, which involve inquiries into digital aspects. Today, to comprehensively cover news in the fairest and most accurate way possible, we can't solely be on physical terrain. We also need to be present in the virtual realm. For instance, since the beginning of the war in Ukraine, over 900 fact-checks have been published by our specialized unit, with support from AFP journalists on the ground. Ultimately, fact-checking is just one branch of digital investigation. All the methods our journalists use to distinguish truth from falsehood are online investigative techniques: verifying the source of a photo or video, researching across time, finding previous versions of a website... These investigative methods can serve more broadly in everyday journalism that isn't about debunking falsehoods but rather reporting the truth.
Amidst this flood of false information, how do you choose the content to fact-check?
A journalist has an area of expertise: economics, politics, sports... They can't cover everything; choices must be made. Similarly, faced with the dissemination of false information, we can't fact-check every subject deserving scrutiny. We have to make choices. Two criteria are crucial in fact-checking, in my opinion. The first: is the content viral? If a piece of false information is shared by three people, we won't give it visibility it doesn't deserve. Virality matters, but it's not absolute. Sometimes, we notice a weak signal a piece of content that isn't viral yet but deserves verification to prevent it from spreading. The second important point is the type of information. There's a wide range of false information. The classic? Massively shared false information that's relatively harmless; for example, during the health crisis when internet users joked about 'The Simpsons' predicting COVID-19 years earlier. That's false. However, other false information can be more dangerous despite having only a hundred shares. But precisely because they can be dangerous, fact-checking these contents is interesting. False information varies in severity and can have more or less real-life consequences. Our choice is primarily determined by the content's virality, whether it poses a danger to individuals, or if it pollutes public discourse.
You mention that these 'investigative methods can be useful in everyday journalism.' What impact does fact-checking have on journalists' work?
Fact-checking initiates changes. Often, journalists have the illusion of understanding the digital realm. It's just an illusion. Few truly master it. Journalists have Twitter, Facebook, or Instagram accounts, but we realize they don't efficiently search on a search engine. On Google, one can conduct extremely precise searches: setting chronological limits, filtering results by file type, excluding a word from the query... 90% of journalists are unaware of these features. That's why we organize internal training at AFP and also externally, teaching search techniques, especially various filters. These methods save time and find more relevant content. In this regard, the digital investigative techniques we've developed in fact-checking can be useful for the entire editorial team.
There's been an increase in false information with the rise of social media. While they contribute to spreading false information, can they also be useful in fighting misinformation?
Today, social media platforms are aware that false information pollutes their ecosystem. While platforms react to varying degrees, they all understand they need journalists to produce fact-checking. If we aim to combat misinformation spreading on these platforms, we must be present there too. That's why AFP currently collaborates with Meta in a program called Third-Party Fact-Checking, along with more than 60 other media outlets worldwide. The process is as follows: if we detect false information on Facebook, we write our fact-check; the platform provides us with a tool to label that publication. The user who shared, reposted, or sees the false information in their news feed will see a warning with a link leading to our verification. This program with Facebook also helps reduce the virality of such content. False information is extremely viral. Once we detect a publication conveying false information, Facebook's algorithm stops giving it visibility. A page repeatedly sharing false information will eventually face penalties, meaning less visibility, inability to advertise, or even exclusion from the platform. The effect is tangible.
Today, some want to 'fact-check the fact-checkers': how do you perceive this?
Generally, those aiming to fact-check the fact-checkers believe our verification isn't convincing. They're free to fact-check us again. But it's not so much AFP that needs fact-checking; it's more about the experts we quote. I'm not saying we're flawless. Like any journalist, we can make mistakes, but we rectify them. If you visit the Factuel blog, we have a page with all our corrections compiled transparently.
AFP is launching an online training to assist all journalists in enhancing their digital investigative skills and combating misinformation. Find the courses here.
Interview conducted by Emma Alcaraz
