The 2016 US elections have opened a very important debate on the other side of the Atlantic: What is the impact of the internet giants at the ballot box? At the centre of the debate is the massive publication of "fake news" or "misinformation" set up in Russia and particularly used and relayed by many supporters of Donald Trump on Google, Twitter and Facebook. Is the phenomenon being replicated for the next elections? NY Times ?
Sundar Pichai, CEO of Google made no secret in an interview with the BBC some flaws in the last US election and thus admitted that "bad decisions were made" and said "this is a learning moment for us" [in content moderation]. He finished by adding "From our point of view, false information should not be distributed, so we all agree that things need to improve".
On the Facebook side, Mark Zuckerberg, its founder, had a little more trouble admitting his mistakes. Although he expressed a willingness to make an effort in moderation, he also added "Identifying the truth is complicated. While some hoaxes can be completely dismantled, a lot of content, including from mainstream sources often has the right basic idea but with false or omitted details."
But what have been the latest major updates from Google, Facebook or Twitter since 2016?
Google's hunt for deceptive doctored images
Here is the type of footage we could find following the 2016 election:
"Donald Trump won a landslide victory by 7.5 million votes in 3,084 of the 3,141 counties or county equivalents in the country."
In order to offer more visibility on the distinction between fake and fact news, Google has developed a "Fact-checking" option to highlight whether a photo or an article has been approved as true or not:
The principle is quite simple: with the help of the structured data ClaimReview third parties will be able to validate or not validate an article or an image presented on Google.
Although Google has announced that this will have no impact on the ranking of images or articles, it will be a significant factor in differentiating between correct and false information.
In the event of a dispute, it is also possible for a site to carry Google claim who will examine the merits of Fact cheking.
Unambitious Facebook against misinformation
In 2016, Mark Zuckerberg, founder and CEO of the world's first social network, had to justify to the US Congress the publication of Russian political advertisements in favour of candidate Donald Trump.
Following a press release published on his personal page However, on 3 September, Facebook's CEO did not announce any major breakthroughs in this area. Facebook's strategy would be mainly to bring as many people as possible to the polls by means of registration links, latest updates and tutorial videos displayed on all the American users' pages.
In terms of misinformation, however, few major actions are worth noting. All the actions have the primary aim of encouraging people to come and vote by avoiding misinformation on the subject (risks of contamination, misinformation on how to vote, etc.). In addition, it will no longer be possible to broadcast new political announcements during the last week before the vote, however all subsequent announcements will be. No verification of these announcements has been announced by the firm. False information concerning the way to vote will be filtered out only 72 hours before the elections.
No filtering of published and shared information has been announced outside of Messenger "We will limit the risk of making misinformation and harmful information viral by limiting sharing on Messenger. It will still be possible to share election information, but you will be limited on the number of chats you share it on". In concrete terms, a message can only be transferred five times on Messenger and even only once for more "sensitive" content. Already implemented on Whatsapp, this is applied to all information related to politics, true or false. We can still deplore no real action to ban fake news from the social network.
Whatsapp on its side has developed a button that allows you to search quickly from a shared item on its platform. When a message is shared multiple times on its platform, a magnifying glass button will automatically be visible on the right of the preview. The preview will allow you to search the web to verify this information. Since it is intended to be simple to use, it will encourage users of the messaging system, which is the property of Facebook, to check whether or not a piece of information is true or not.
This option is currently only available in the United States, Brazil, Spain, Ireland, Italy, Mexico and the United Kingdom.
Twitter and fake news?
Twitter made headlines a few weeks ago when it masked a tweet from Donald Trump that turned out to be false information about the postal vote.
Among its actions, we can mention the deletion of many accounts. In particular, we could find accounts masquerading as "Black lives matter" protesters and subsequently wishing to vote for Donald Trump.
We have in this example many false accounts sharing the same discourse "I've been a Democrat all my life. I joined the Black Lives Matter protest in the beginning. They opened my eyes. I hadn't noticed that I had become a Marxist. It happened without me even realizing it. I'm done with that bullshit. I'm going to join the Republicans. Give Twitter a chance."
This message, relayed by dozens of accounts, prompted Twitter to take action to delete the accounts that were copying and pasting the message.
The American elections of 2020 in the image of 2016?
The question is therefore pending for the time being. There is no doubt today about the impact of social networks on elections. The major IT companies have shown this by signing together last month a commitment to "protect integrity" in the upcoming elections. Among these participants were Facebook, Google, Twitter, Reddit, Microsofit, Pinterest, LinkedIn, Wikimedia and Verizon Media. We have now seen that everyone has adopted different resolutions and we can still wonder whether all these actions will succeed in limiting the spread of misinformation. To be continued!