Fake news, real threat

Manipulating social media is easy with no downside for failure

The US presidential elections represented a watershed moment for social media, which has often been used as a vehicle for political messaging. But this is the first time it is being seen as a malign platform for dissemination of fake news. Facebook, in particular, was inundated with fake political news that many suspect had a major influence on the elections. Many Americans use social media as a source of information, accessing news via Facebook and Twitter. According to Pew Research, Facebook is the primary news source for 44% of American adults; 9% use Twitter. But sourcing news only via social media leaves the user vulnerable to fakery.

 
For instance, over 100 fake news sites favouring Donald Trump were located on servers in Macedonia and spewed fictions and insinuations such as the Pope endorsing Mr Trump and Hillary Clinton arranging the murder an FBI agent investigating Wikileaks emails. It is easy to disseminate fake news on social media by setting up websites in faraway places. But it is very hard to sift lies and slanted half-truths from the unvarnished truth without recourse to mainstream news sources. Such fake news was promoted on social media platforms and propagated through repeated sharing by bogus, automated accounts. Somewhere between 5.5 and 11% of Facebook’s 1.8 billion accounts are estimated to be fake. Even the Twitter followers of Mr Trump (13 million) and Ms Clinton (10 million) contain many bots — although Republican Twitter bots outnumbered Democratic ones by a ratio of 5:1 on election day.

 
Mainstream media can flag such absurd stories and this facility is also available via Facebook. But users set up “filter bubbles”, which ignore the mainstream media. Facebook users choose their friends and user-groups and click “like” on posts and pages. Twitter users “follow” accounts they choose. This creates confirmation bias and the resulting bubble excludes news and opinion that does not conform to the user's worldview. Making things worse, Facebook amplifies the filter bubble effect by its secret algorithm, which delivers a filtered feed, including promoted news and ads. The algorithm analyses usage patterns, “likes”, comments and group memberships to create unique feeds for every user. The priority is to create engagement; not to check for credibility.

 

Digital marketers use Facebook's analytics and advertising tools to tailor campaigns to target specific demographics. Facebook does not reveal how much revenue it earns from news promotion or political advertising and although CEO Mark Zuckerberg pooh-poohed Facebook's political influence, the network's own marketing highlights “success stories” in influencing “government and politics”. For example, it claims to have influenced the Arab Spring and helped the Scottish National Party. Google also allowed fake news sites to generate advertising revenue as partners to its AdSense programme and its search algorithm, PageRank, does value credibility and authenticity.
 
To be sure, there is increasing awareness about the issue. A group of students recently demonstrated that it was possible to build automated tools to verify news on Facebook. For its part, Facebook has said it will henceforth filter out “misinformation” with Mr Zuckerberg outlining the strategy in a detailed response. Google says it is now committed to denying AdSense partnership to fake news sites.

 
However, even as the platforms try to flag fakery, purveyors of phoney news will try to game the system. The rewards for successful manipulation are high and there is no downside for failure. It is also unrealistic to expect social media platforms to verify the news if users live in a bubble and abdicate their own responsibility to check for veracity. Mainstream media must also find ways to reassert itself in this strange new world where a single extremist sitting in a basement can match global news services in terms of reach.