First Draft MD Jenni Sargent spoke at the FIPP’s Digital Innovators’ Summit (DIS) in Berlin in March, where delegates rated he talk as one of the top presentations at the summit. Read an edited version of her talk below, or if you prefer, watch the video here instead.
You would assume that by now every newsroom in the world should have skilled journalists who are able to use social media content and online skills to determine if stories are fake or not, says Jenni. “But we find that these skills that should be taught in every journalism school in the world are lacking.”
To cross this void, nine “well meaning individuals” created First Draft in 2015. Today this coalition has 100 partners fighting online misinformation “for the greater good”. The good news is that change is taking place. Today, in some newsrooms there are entire teams dedicated to social search and social verification of stories.
Yet, warns Sargent, the processes for social verification of a story differs radically from those established in traditional newsrooms. “In a perfect world we should be able to pick up a telephone to make the conclusion if a particular story is true of false.” Online, despite many free and easy-to-use tools that should help to investigate the source of content, it remains complicated. On top of this there are also ethical concerns. To better assist responsible news organisations First Draft has been working and collaborating with other like-minded organisations to create tools that will simplify verification processes to identify fake news and prevent the spread of it.
One of these projects in collaboration with the Public Data Lab network was to set up and facilitate research, public engagement and debate to create A field guide to fake news.
The collaboration sets out to “map the misinformation ecosystem” within the context of European elections. Researchers set out to gage patterns of behaviour for fake news generation. The guide should provide answers to questions such as:
- Is the news emerging from one place of source?
- Are there aspects to specific stories that can be identified that will raise suspicion of the authenticity of those stories?
- Can you identify fake stories by identifying the same kind of headline wording?
- Are certain website domains spreading fake news registered to specific locations?
Sargent says by asking and applying these questions a tool such as ‘Mentionmapp’ could discover that 26 Twitter accounts microblogging during the run up to the Dutch elections earlier this year were using the same words and hashtags to spread misinformation. “Those 26 (fake) accounts had a major reach and showed a coordinated effort (to spread misinformation).”
To truly understand the complexity of fighting fake news, says Sargent, those who are serious about combating it, need to understand the difference between misinformation and disinformation. “Misinformation is information that is false but the persons who disseminate it believe that it is true. The major problem here is that this is your average social media user. They are not deliberately sharing something that is false because they want to manipulate the news agenda and disrupt a voting process. They are trying to be helpful by sharing something they find interesting.”
Disinformation, on the other hand, is “a very coordinated, deliberate effort. They know that they are spreading something false and they are deliberately doing so… and they are using tried and tested techniques to spread disinformation.”
To make sense of the complexities of all of this, First Draft has come up with a ‘scale of intent’ to gage those responsible for creating fake news. On the one end there are satire and parody websites whose intent are not to spread fake news but entertain. On the other end are those who create 100 per cent “terrifyingly” false content. In between there is a grey area which comprises bad reporting, lazy sourcing and inept crediting.
To assist newsrooms to identify parody and false content, First Draft develops easy-to-use checklist and tools. One other solution First Draft is experimenting with, “although it is not perfect yet”, is visual icons. Similar to the blue like icon on Facebook, First Draft is starting to ask social platforms to put “satire icons” or “fabricated icons” on content. “There are now ways that we can very quickly identify if something is false. So how do we help audiences and users to see that it is false… without having to ask social platforms to remove it completely?” Visual icons might be one way to start the process, suggests Sargent.
She also says with the right tools it is really quite simple to identify fake websites, imposter websites, fake headlines and fake content. “The challenge is whether we can expect others (users) to do those checks. Or is it our responsibility as publishers to (collectively) do those checks?”
As a first initiative to collectively share responsibility between publishers to combat the distribution of fake news, First Draft launched ‘Cross Check’, which started as “a seed of an idea” last year to prevent the spread of fake news in the run up to the French general election. The initiative now has 37 newsroom partners who work together to cross check every single report they receive and verify that they are willing to endorse the authenticity of it.
As soon as all the newsrooms have endorsed a report, it can be confidently shared on social media and also carries a visual icon to confirm the fact that all the news organisations are endorsing its validity.
“This helps to restore the public’s level of trust in the media and, if anything else, it is helping users to know that the media are working together to fight the misinformation war,” says Sargent.
***Join us for our next event, the 41st FIPP World Congress, taking place in London from 9-11 October 2017. Discounted, pre-agenda tickets are available until 30 April, with savings of £800 on final rates. Learn more and register at fippcongress.com***
Get stories like these delivered directly to your inbox by subscribing to our weekly FIPP World newsletter – sign up here.
More like this