As South Africa goes through its most competitive election campaign ahead of the 8 May 2019 election, since the first all-race democratic elections in 1994, the viral spread of disinformation through social media is now become a real threat to the democracy.
South Africa is a typical developing country, where the youth form the largest segment of the population, as well as the largest voting bloc. Many young increasingly get their information from online platforms, such as Facebook, Twitter and Whatsapp. This means that disinformation on social media could have a significant impact on voting patterns.
In South Africa there has been a proliferation of fake social media profiles and websites, which looks very authentic, but which attack public figures, push racially divisive views and extreme populist policies. These fake social media accounts are both human and robots. Many of them have foreign Internet Protocol (IP) – the numeric label assigned devices connected to a computer network that uses the internet - addresses.
These disinformation peddlers turn marginal issues into the mainstream, aim to create conflicts where there aren’t and distort the focus of public priorities away from critical areas. The most publicised incidents of online disinformation were during the US 2016 elections when it was alleged that a group linked to the Russian government, called the Internet Research Agency, created faked online profiles, which spread fake news about the candidates.
In one occasion, in May 2016 a rogue group, using bots, is alleged to have simultaneously organised fake anti-Islam and pro-Islam in front of an Islamic cultural centre in Houston, to create chaos. The Russian has vehemently denied being involved in the spread of online disinformation.
Ahead of the Canadian federal elections in 2011, misleading robocalls were made to voters directing them to the wrong polling stations. Elections Canada, the country elections organising agency ahead of Canada’s October 2019 federal elections will be using low-tech paper ballots for voting, which is deemed more cyber-attack proof.
In September 2018 Facebook, Google and Twitter signed a voluntary code of practice with the EU in which they pledge to tackle online disinformation campaigns on a voluntary basis. Signatories to the code committed to close accounts and websites that spread disinformation, making it easier for consumers to lodge complaints, and improving the visibility of credible content.
In January this year, Facebook, announced it has toughened political advertising rules and tools in major countries holding elections in 2019 to prevent online election manipulation. The Facebook platform has been used by rogue actors to spread fake news, spread untruths about opponents and cause divisions between communities.
It promised that it will verify the locations and identities of buyers of online political adverts in key country elections taking place in 2019. For example, during the Nigerian presidential elections in February 2019, Facebook only allowed online advertisers to run electoral adverts if they are located in the country. During the Indian general elections currently taking place (11 April to 19 May 2019), Facebook has placed electoral ads in a searchable online library for seven years.
The Indian archive will have the information of ad buyers or the proof of their legitimacy. Facebook will verify whether the names of the buyers of the political ads correspond to those who got government approval to advertise. In Australia, Twitter have implemented new rules that all political ads show who sponsored them; and identify whether they are located in Australia. Facebook in Australia proposed that users posting political ads must verify their locations.
Facebook will introduce similar safeguards in the European Union parliamentary elections later in May 2019. Last year, Facebook introduced searchable online libraries for electoral ads in the US, Britain and Brazil.
In January 2019 the YouTube, said it will change the structure of algorithms to recommend fewer conspiracy theory and extremism videos. YouTube has a recommendations feature, which automatically recommends a playlist for viewers to watch. The recommended playlist is based on algorithms that calculates a viewer’s preferences based on past uses. The YouTube function has been criticised for recommending users to harmful, fake and extremist content, the users would not have otherwise viewed.
The distributers of online fake news target people with sham news based on people’s specific information preferences. They found out peoples online preferences through using social media analytics, the activity through which data is gathered from one’s conversations, engagements and interactions on digital platforms, and arranged these into patterns or trends.
Such analytics work through the process of cookies, which are files websites place on a computer to save the users online preferences, searches and sites visited. When a website a user visits install a cookie to the user web browser or application, the cookie assign a unique ID to the user.
When you visit the website again, the cookie saved in your browser, now assigned with your unique user ID, will identify you, and tell the website to use your previous preferences. The cookies also monitor your activity to all other websites you visit. Many websites have “tracking” cookies which track that your browser through its unique ID, have visited them.
Trackers from companies, such as Facebook, Google and other companies, track your web visits – and are then able to analyse your online preferences, and will send you personalised information, news and adverts. Models which can predict your interests, purchases and news likings, can then be build based on your user preferences. Your preferences can then be sold to advertisers.
Facebook have algorithms to calculate which messages to send to whom based on their o online preferences. The phenomenon of “micro-targeting” describes the method of sending messages specifically based on uses preferences. The data harvesting of users has rightly been criticised. Most people are unaware their data are being mined. Hackers have frequently accessed such mined personal data of users, and published it online.
Social media organisations should extent the EU-like agreements to police online fake news to countries such as South Africa. Social media platforms should close down fake accounts, whether human or from robots that spread fake news. South African civil society organisations will have to monitor potential online disinformation, fake news and smear campaigns better. Citizens must be more discerning about websites they visit and social media content they read or watch.
William Gumede is Executive Chairman, Democracy Works Foundation (www.democracyworksfoundation.org) and author of South Africa in BRICS. (Tafelberg).