Log into your member account to listen to this article. Not a member? Join the herd.

Following the tabling of the Election (Amendment) Bill 2022 in the National Assembly by the Majority Leader in February this year, there was significant debate and angst about the Bill’s proposed amendments to the provisions in the Election Act relating to the transmission of results. Several news outlets and social media users incorrectly asserted that the Bill proposed to scrap the Independent Electoral and Boundaries Commission’s (IEBC) obligation to set up and operate an electronic tallying and transmission system, reverting Kenya’s elections procedure to that which was last used in 2007. The traction gained by these assertions prompted the IEBC to clarify the position via a press statement—the provision relating to electronic transmission was not going to be amended.

This episode is emblematic of two things: an apparent recognition of the importance of electronic transmission and tallying; and an insight into the potential for social media to muddy the waters of public discourse in electoral contexts. More broadly, it is symbolic of the way electioneering—and the administration of elections—has drastically evolved in recent years. The level of digitisation of both the election administration systems and the tools used by candidates to campaign is unlike anything previously witnessed, and this is a mixed bag. While the use of technology may make processes more efficient and secure, there are inherent risks posed both to electoral systems, and to the electorate itself. In a two-part article within this series, we canvas some of the key challenges posed by the digitisation of electoral processes and review some of the measures adopted by various stakeholders in response. We look at these challenges at two levels—the structural risks posed to election administration systems and the risks posed to electoral integrity by conduct on social networks.

Election administration systems

Several election management bodies have begun using technology to administer elections to varying degrees. While some such as Estonia have implemented a system which allows citizens to cast their vote online from anywhere in the world, others, such as Kenya, still limit their use of technology to voter registration, transmission, and tallying. While these systems may render numerous efficiencies, this is sometimes at the cost of credibility. Any challenge to the integrity of these systems would effectively undermine entire electoral processes. One need only look at Kenya’s most recent experience to understand this.

After the 2017 general election, the Supreme Court of Kenya ordered the IEBC to allow two independent parties access to its servers to audit the results of the poll. This came after allegations that there were significant flaws in the tallying and transmission of results. During the various election related petitions which the Supreme Court heard in that period, public discourse around the reliability of the electronic tallying and transmission system demonstrated widespread mistrust. Considering the Supreme Court subsequently accepted the claim that the data may have been interfered with, this was unsurprising. The ambivalence towards the reliability of electronic tallying and transmission was apparent prior to the election, when officials struggled to instil confidence in voters that the systems would be reliable. In the face of clear hurdles such as insufficient network coverage, the IEBC had a difficult time assuring the electorate that it would be able to safeguard an election. Since then, there have been several instances where the integrity of the IEBC’s systems has been questioned. For example, only last year, there were conflicting reports from the Directorate of Criminal Investigations (DCI) and the IEBC on an alleged infiltration of the IEBC’s servers. While the DCI publicly stated that an individual had been arrested on suspicion of having hacked into the IEBC servers, the IEBC denied these claims.

The level of digitisation of both the election administration systems and the tools used by candidates to campaign is unlike anything previously witnessed.

With digitised election administration, election management bodies must consider both internal and external risks. From flaws inherent in the systems used, to the potential for tampering by external actors, there exist numerous vectors through which a system’s reliability can be challenged. Once this reliability is impeached, it is hard to guarantee the credibility of any result. With this reality, it is now common for various political actors around the world to allege external interference or inherent unreliability when confronted with an unfavourable outcome. In countries where this occurs, the resulting mistrust of election institutions threatens the very fabric of democracy, sometimes leading to violence. Lest we forget, this is not a developing world phenomenon. The United States of America experienced this in light of Donald Trump’s rejection of the US Presidential elections.

Social networking platforms

Beyond the cybersecurity threats posed to election administration systems, the digitisation of electioneering has also exposed democracies to several harms. Over the past few election cycles, the use of networked technologies—social media in particular—in political campaigning has steadily risen. During the 2017 elections, over 80 per cent of candidates in Kenya had an online presence on social media. The peer-to-peer and instantaneous nature of these platforms means that they are especially potent tools for political actors to use for the dissemination of their messaging to large audiences, as well as for direct engagement with individual members of their bases. Given that these platforms often lack gatekeeping in the traditional editorial sense, political actors can share unfiltered content with large audiences easily.

On the one hand, these platforms have been lauded as levelling the playing field between actors who have access to the resources necessary to command a presence on mainstream media and those who don’t. On the other, the way these platforms curate the content their users consume, and the nature of the content shared by some political actors, have been criticised as undermining healthy political debate. Specifically, the spread of harmful content such as mis- and dis-information which contributes to the erosion of the basic truths which underpin political discourse and is sometimes inciteful has been of significant concern. Public authorities and private entities alike are increasingly finding themselves having to issue statements disavowing the contents of news reports or social media posts as false or misleading, as the IEBC did in relation to the reporting following the tabling of the Election (Amendment) Bill, 2022.

The same is true in relation to announcements claiming to be from public bodies.  The quality of the fakes is excellent so the public can be easily deceived.  Advances in artificial intelligence (AI) also mean that audio-visual content is not immune to these harms; the democratisation of AI technologies used to alter audio-visual content has resulted in a surge in fabricated videos known as “deep fakes”. With such an information ecosystem, it becomes difficult for political actors and the electorate to engage in constructive debate. Immediately one is confronted with incriminating content such as a video, one need only claim that it is fabricated, and this will immediately obfuscate the truth. Focus will be shifted from the content of the post to its authenticity. While social media platforms have developed some fact-checking and transparency measures in relation to, for example, the US elections, many have not bothered to do the same for most of the rest of the world, which itself is a matter of significant concern.

With digitised election administration, election management bodies must consider both internal and external risks.

These harms are compounded by practices such as microtargeting, which leverage the design of social media platforms to expose specific audiences to messaging which appeals to their biases, and which they would therefore be receptive to. In campaign contexts, these tools are often made available to political actors by consultants such as the now infamous Cambridge Analytica which offered its services to various actors in Kenya’s 2017 elections. Underpinning these targeting and microtargeting practices is an aggressive and often unchecked harvesting of personal data which enables political actors to develop audience profiles and design effective messaging. In some instances, this messaging leverages on an audience’s biases or fears, and is sometimes misleading or inciteful in nature. Recently we have also witnessed the use of social media influencers to coordinate the spread of this specific messaging that is often false, misleading, or maligning of a particular political actor or group of actors. Social media platforms, in response to findings by researchers, have severally had to take down a raft of accounts engaged in coordinated inauthentic behaviour. Most recently, this method of spreading mis-and dis-information was used in relation to the debate around the Reproductive Healthcare Bill and the Surrogacy Bill where a Spanish organisation—CitizenGo—allegedly amplified certain hashtags on Kenya’s trending page on Twitter.

Vigilance is then the watchword. Building “believability” filters and inculcating healthy scepticism are essential. In years past, the fact that something made it into print media anointed the content with a sense of credibility.  Social media has irrevocably changed this. These developments mean that election stakeholders ought to be vigilant regarding the ways in which Kenya’s electoral integrity can be attacked as we head into the 2022 general elections. On the other hand, these challenges have been widely noted by the IEBC, other regulators such as the Office of the Data Protection Commissioner, and civil society, and various measures have been put in place to safeguard Kenya’s electoral integrity. We discuss some of these measures in the next instalment in this series.

This article was authored in collaboration with the Kofi Annan Foundation whose electoral integrity program is supported by the United Nations Democracy Fund.