13 - Navigating the Nexus of Elections, Technology and Democracy amid Escalating Disinformation and Misinformation Challenges in Kenya

Summary

Despite Kenya’s transformative and progressive 2010 Constitution, it is still grappling with a hybrid democracy, displaying both authoritarian and democratic traits. Scholars attribute this status to several factors, with a prominent one being the domination of the political order and wielding of political power by a few individuals and families with historical ties to patronage networks and informal power structures. The persisting issues of electoral fraud, widespread corruption, media harassment, weak rule of law and governance challenges further contribute to the hybrid democracy status. While the 2010 Constitution aims to restructure the state and enhance democratic institutions, the transition process is considered incomplete, especially since the judiciary’s role of judicial review is mostly faced with the difficult task of countering democratic regression. Moreover, critical institutions such as the Independent Electoral and Boundaries Commission (IEBC) have faced criticism due to corruption scandals and perceptions of partisanship, eroding public trust in their ability to oversee fair elections effectively.

13.1 Introduction

Despite Kenya’s transformative and progressive 2010 Constitution, it is still grappling with a hybrid democracy, displaying both authoritarian and democratic traits. Scholars attribute this status to several factors, with a prominent one being the domination of the political order and wielding of political power by a few individuals and families with historical ties to patronage networks and informal power structures.¹ The persisting issues of electoral fraud, widespread corruption, media harassment, weak rule of law and governance challenges further contribute to the hybrid democracy status.² While the 2010 Constitution aims to restructure the state and enhance democratic institutions, the transition process is considered incomplete, especially since the judiciary’s role of judicial review is mostly faced with the difficult task of countering democratic regression.³ Moreover, critical institutions such as the Independent Electoral and Boundaries Commission (IEBC) have faced criticism due to corruption scandals and perceptions of partisanship, eroding public trust in their ability to oversee fair elections effectively.⁴

In the context of Kenya’s hybrid democracy, the new challenge posed by disinformation and misinformation has become a pressing concern, particularly in relation to elections. The selective use of information to manipulate political outcomes has led to a detrimental impact on the democratic process, contributing to issues such as voter apathy, ethnic polarisation and a sense of disorientation caused by the rise of misinformation. As this chapter highlights, technology can indeed serve as a tool to enhance democratic legitimacy and combat election rigging when properly implemented and managed. However, it is crucial to recognise that technology, especially digital information technology and social media, has become a double-edged sword. On the one hand, these platforms play a significant role in bolstering democracy by facilitating broader public participation, a fundamental aspect for any democratic society to thrive. Yet, on the other hand, the increased accessibility of the Internet has given rise to a new challenge: the spread of disinformation and misinformation.⁵

This phenomenon poses a serious threat to the integrity of elections and democratic processes. As the overwhelming levels of disinformation and misinformation in Kenya’s 2022 General Election illustrate, misleading information can easily circulate, influencing public opinion and undermining the trust that citizens place in democratic institutions. The proliferation of false narratives skews the democratic playing field, making it difficult for voters to make informed decisions based on accurate and reliable information. Therefore, it is essential for the government, the election management body, civil society, tech companies and other stakeholders to tackle this issue collaboratively. As this chapter discusses, striking the right balance between the benefits of technology in enhancing democracy and countering its potentially negative effects requires concerted efforts to promote fact-checking, media literacy and responsible digital citizenship. By equipping citizens with the skills to distinguish accurate information from disinformation and misinformation, Kenya can safeguard the democratic ideals upon which the society is built and ensure that technology remains a tool for progress rather than a source of division. As Kenya prepares for future elections, addressing the challenges posed by disinformation and misinformation and their impact on the democratic process will be crucial in preserving the fairness and credibility of elections.

This chapter is organised as follows: Section 13.2 delves into the historical context of Kenya’s hybrid democracy and analyses the prevailing challenges in the electoral landscape. It investigates the complexities arising from ethnic politics and election rigging that have hindered the democratic process. Section 13.3 examines the role of technology in enhancing electoral integrity and legitimacy, particularly in relation to the role of technology in Kenya’s 2013 and 2017 general elections. It explores how the IEBC endeavours to leverage technology to administer elections effectively, aiming to curb election illegalities and irregularities. Section 13.4 sheds light on the growing challenge of disinformation and misinformation, particularly during the 2022 general election. It examines the weaponisation of fake news by politicians seeking electoral advantages, and the issues around technology and reliance on false information that arose in the 2022 presidential election petition. Section 13.5 looks at legal, institutional and non-legal measures for detecting and combating disinformation and misinformation. It discusses the pivotal roles various stakeholders such as the IEBC, tech companies and civil society actors play in detecting and combating false information. Further, it emphasises the importance of empowering citizens with media literacy and critical thinking skills to enable individuals to distinguish accurate information from misinformation. Section 13.6 then concludes. It offers a recapitulation of the findings and implications for Kenya’s electoral system.

13.2 Laying the Foundation: History of Kenya’s Hybrid Democracy and Challenges in the Electoral Landscape

The high stakes and divisiveness of elections is at the centre of Kenya’s woes and resultant cycle of ethnic voting and violence that follows every election.⁶ Ethnic politics in Kenya dates back as early as the pre-colonial era, to colonisation and upon independence. The British colonisation of Kenya employed a ‘divide and rule’ policy that played ethnic groups against each other to prevent a unified nationalist opposition. This strategy was later adopted by post-colonial political elites who saw their ethnic group as their crucial asset when negotiating for ‘a seat at the national table’.⁷ At the time of independence, Kenya, like many other newly independent African states, prioritised nation-building and socio-economic development. With a limited number of schools, scarcity of hospitals and inadequate infrastructure, the political leaders of the era directed their efforts towards addressing pressing challenges such as high levels of illiteracy, poverty and disease.⁸

The authoritarian model, which provided the President and executive with significant authority to regulate the allocation and distribution of resources, was seen as best to tackle these issues.⁹ However, to remove it from its colonial origins to fit the post-colonial agenda, its mission was changed and repackaged as being necessitated by the ‘development first’ agenda:¹⁰ to win the war against poverty, ignorance and disease, ‘the development plan trumped the Constitution as the most important economic and political document of the state’ and hence a ‘worthy trade-off’.¹¹ In addition, one-party politics was preferred to multi-party politics due to the fear that the latter would encourage partisan mobilisation of subnational identities and loyalties manipulated by colonialism, ultimately undermining national unity and hindering the post-colonial development project’s success.¹² However, despite the ambitious development project, several challenges arose, including rampant corruption, stagnant or negative growth rates, worsening income inequality and a rise in absolute poverty.¹³ As a result, the ‘development first’ initiative ultimately proved unsuccessful.

As public trust in the ruling elite waned, a troubling practice used during the colonial era re-emerged, where certain ethnically aligned sections of the citizenry, particularly those offering strong political support to the ruling elite, received preferential treatment and benefits in exchange for their loyalty.¹⁴ This strategy was used by the ruling elite to maintain their grip on power and retain political support. It marked the inception of divisive ethnic politics in independent Kenya, characterised by election rigging through means such as ballot box stuffing, state-sponsored ethnic violence and a blatant disregard for the rule of law. Incumbents resorted to desperate measures to cling to power, turning elections into a zero-sum ‘winner takes all’ game. Those in control of the political order wielded significant influence, becoming the primary dispensers of favours and socio-economic benefits, leading to a system where loyalty to the ruling party was rewarded while dissenting voices faced marginalisation.¹⁵ The emergence of ethnicised politics gave rise to profound and enduring ethno-regional inequalities, as politically dissident ethnic groups and their respective regions faced socio-economic marginalisation.¹⁶ It also fuelled ethnic hatred and divisions.

Compounding this problem, a series of amendments to the 1963 Constitution exacerbated the establishment of an authoritarian governmental structure, concentrating power within the presidency and the executive under one-party rule.¹⁷ This concentration of power disrupted the equilibrium between the executive and legislative branches, severely curtailing parliament’s ability to effectively supervise the actions of the executive.¹⁸ Concurrently, the independence of the judiciary also suffered. Particularly noteworthy was the removal of judicial security of tenure for High Court and Court of Appeal judges in 1988, which rendered judges susceptible to removal at the discretion of the President.¹⁹ Although this amendment was later repealed, its initial intention was to further consolidate executive power. The amendment weakened the judiciary’s credibility, subjected judges to the executive’s whims, and eroded the essential ‘safeguards necessary for maintaining fair administration, neutrality of public institutions, accountability of government, and the protection of rights in general’.²⁰ Consequently, the essential system of checks and balances necessary to uphold democracy within the three governmental branches was largely deficient, leading to adverse implications for the overall preservation of democratic principles.

The push for electoral reforms in Kenya took root in the 1990s when there was a growing demand for the repeal of section 2A from the 1963 independence Constitution. This section, introduced through a constitutional amendment in 1982, had transformed Kenya from a de facto to a de jure one-party state, leaving only the ruling party, the Kenya African National Union (KANU), in existence and outlawing political opposition.²¹ This meant that individuals could only vie for political office if they were members of and nominated by KANU. The repeal of section 2A in 1991 and the subsequent restoration of multi-party politics was a significant milestone.²² However, it soon became evident to the reform movement that this alone was ‘insufficient to democratise politics, usher in accountability and ensure responsive political policies’.²³ The 1963 Constitution, along with its various amendments, was deemed undemocratic, solidifying an authoritarian system of government, which made it difficult to challenge the ruling party’s political power.²⁴

The electoral landscape remained contentious, marred by ethnic hate-speech, violence and persistent allegations of rigging. This environment provided incumbents with a clear advantage, undermining the democratic process, and perpetuating their stay in power.²⁵ One of the key contributing factors to this situation was the partisan appointment of commissioners to the Electoral Commission of Kenya (ECK), which eroded public trust and raised questions about the Commission’s competence.²⁶ As a result of these challenges, there was an urgent need to establish mechanisms that would ensure transparent, inclusive, and credible elections. This imperative became a central agenda in the pursuit of constitutional reforms. The aim was to foster a system that could guarantee the integrity of the electoral process, promote fair representation, and restore public confidence in the democratic foundations of the country. The ECK made efforts to introduce technology before the 2002 and 2007 elections to support transparent and efficient administration of election processes, but it was generally slow in adopting the use of technology.²⁷ There was no clear strategy for technology integration and the process lacked statutory and regulatory backing. Moreover, stakeholders were not consulted, and staff were not properly trained on the use of technology in election administration.²⁸ The result was that technology had limited impact in results management, and a failure to report preliminary results in the December 2007 elections contributed to the widespread suspicion that the ECK lacked transparency and possibly manipulated the results.

After facing several challenges and false starts in the journey towards constitutional reform,²⁹ Kenya finally achieved a major milestone with the promulgation of a new constitution in 2010. This significant achievement followed a constitutional referendum, in which an overwhelming 67 per cent of Kenyans voted in favour of the Constitution. The success of the referendum can be attributed, in part, to the profound impact of the devastating 2007/2008 post-election violence that left a death toll of over 1,000 people and thousands more wounded and internally displaced.³⁰ This served as a wake-up call, highlighting the critical need for inclusivity and governance based on the rule of law rather than allegiance to a ruling elite or ethnic favouritism.

To further understand the changes Kenyans sought in the constitutional review process regarding elections, insights from the 2008 Independent Review Commission Report, commonly known as the ‘Kriegler Report’, which emerged following the 2007/2008 post-election violence, are valuable.³¹ The Kriegler Report highlighted severe weaknesses in the ECK, leading to a loss of public confidence and institutional legitimacy. It thus recommended an overhaul of the electoral management process.³² Consequently, the ECK was disbanded and replaced by the Interim Independent Electoral Commission (IIEC) and Interim Independent Boundaries Review Commission (IIBRC). These interim bodies were tasked with revamping the electoral management system, introducing technology in elections, creating a new voters’ register and proposing boundary delimitation reforms.³³ It is the IIEC that successfully conducted the constitutional referendum that led to the enactment of the 2010 Constitution.

13.3 Harnessing Technology for Electoral Integrity and Legitimacy: The Role of Technology in Kenya’s 2013 and 2017 General Elections

The history of Kenya’s elections has been marred by human interference through election rigging, voter bribery, state-sponsored ethnic violence and media blackouts during the announcement of presidential election results.³⁴ These issues continue to exacerbate the divisiveness of elections in Kenya. As highlighted in the previous section, the Kriegler Commission Report noted the reluctance and indifference of the ECK to take steps to integrate technology and develop systems for results transmission and procurement. This Commission considered various proposals, including one from the International Foundation for Electoral Systems, advocating the integration of technology to streamline and facilitate electoral processes.³⁵ While acknowledging the efforts made between the 2002 and 2007 elections, the Kriegler Commission concluded that the introduction of technology in elections was inevitable to address existing challenges and enhance the integrity and legitimacy of the electoral process.³⁶ Notably, scholars perceived the adoption of technology in elections as a means to improve ‘the administration of elections and increase voter confidence because of its perceived levels of accuracy, verifiability and efficiency as compared to manual systems’.³⁷ Consequently, the Kriegler Commission recommended that there be established ‘without delay … an integrated and secure tallying and data transmission system’.³⁸

The aim was to revamp the electoral management system and use technology as a key tool in election management processes, including the creation of a new voters’ register. Subsequently, the IEBC was established upon the promulgation of the 2010 Constitution and took over the mandate of the IIEC and the IIBRC.³⁹ The Independent Electoral and Boundaries Commission Act 2011 was then enacted as the first statute dedicated to the management and conduct of elections, outlining the Commission’s additional functions, including the use of appropriate technology in its activities.⁴⁰

Section 44(1) of the 2011 Elections Act provides for the use of technology in election processes. For the presidential election, the IEBC is required to transmit results electronically from polling stations to the constituency tallying centres (CTCs) and national tallying centres (NTCs) through the Results Transmission System (RTS). In addition, the IEBC is obligated to livestream and maintain a public portal for transparency and accountability.⁴¹ By embracing technology in the electoral process and adhering to regulations that promote accessibility and inclusivity, the IEBC seeks to restore public trust and confidence in the electoral system, fostering a fair and credible electoral process in Kenya.

13.3.1 The 2013 General Election

The Kenyan public’s hope for credible elections was evident in the 2013 general election, where a record 12 million Kenyans, accounting for 85.9 per cent of registered voters, turned out to vote – the highest turnout since the reintroduction of multi-party politics in Kenya.⁴² The election marked the first use of the Biometric Voter Registration (BVR) System, an electronically based national register, replacing the manual register known as the Green Book for voter identification. The BVR System was accompanied by the Electronic Voter Identification System (EVID) to verify voters’ identities.⁴³ The 2013 general election witnessed a closely contested race between two presidential candidates, Uhuru Kenyatta and Raila Odinga. After the election, Uhuru Kenyatta was declared the winner with 50.5 per cent of the votes. However, the outcome was quickly challenged in the consolidated petition of Raila Odinga and 5 Others v. Independent Electoral and Boundaries Commission and 3 Others.⁴⁴

A significant point of contention in the petition was the failure of the RTS and the BVR, which led to multiple variations in the electronic data generated.⁴⁵ In its judgment, the Supreme Court acknowledged the irregularities in data and information capture during the registration process, but deemed them insufficient to affect the credibility of the electoral process.⁴⁶ Despite acknowledging various irregularities, the Court found that the evidence did not demonstrate profound irregularities in the presidential election’s conduct that would invalidate the final result.⁴⁷ Thus, the petition was disallowed, and the Supreme Court upheld the election’s outcome.

In response to the discrepancies identified in the 2013 presidential election petition regarding the use of technology in managing and conducting elections, the Election Laws (Amendment) Act of 2016 was enacted to introduce legislative reforms related to election technology to address the issues. One crucial change was made to Section 39 of the Elections Act, requiring the mandatory electronic transmission of presidential election results from each polling station to both the CTCs and NTCs). The Act also mandated the IEBC to publish polling results forms on a publicly accessible online portal maintained by the Commission.⁴⁸

Another significant amendment was made to Section 44 of the Elections Act, which established an integrated electronic electoral system, the Kenya Integrated Elections Management System (KIEMS). This system was designed to facilitate various functions, including biometric voter registration, electronic voter identification and electronic transmission of election results. To ensure the integrity of the technology used, the amendment stipulated that the electoral system must be simple, accurate, verifiable, secure, accountable and transparent.⁴⁹ These amendments were aimed at enhancing the reliability and credibility of election processes by leveraging technology.

13.3.2 The 2017 General Election

The challenges faced by IEBC in ensuring the proper conduct of elections and curbing illegalities and irregularities during the 2013 general election resulted in a significant decline in voter turnout, from 85.9 per cent in 2013 to 77.5 per cent in the 2017 general election.⁵⁰ This reflects a growing sense of voter apathy. The 2017 general election was held on 8 August 2017, and it was anticipated that the IEBC would address the technological issues encountered in the 2013 election. To achieve this, the IEBC implemented the KIEMS for the first time.⁵¹ While the electoral technology performed better than it did in 2013, technological challenges were still encountered, affecting the perceived credibility and legitimacy of election results.⁵²

Following the declaration of the incumbent president, Uhuru Kenyatta, as the winner by the IEBC returning officer for the presidential election on 11 August 2017, the opposition filed a petition, Raila Amolo Odinga and Another v. Independent Electoral and Boundaries Commission and 2 Others,⁵³ challenging the election’s fairness and legality. Numerous election irregularities and illegalities were cited, particularly in relation to election technology. The petitioners sought an audit and scrutiny of all presidential election forms and the election system, including Forms 34A, 34B, 34C, and the KIEMS kits, server(s) and website/portal.⁵⁴

In relation to the functions of Forms 34A, 34B and 34C, as per section 39(1C) of the Elections Act, upon tallying presidential election results at a polling station, the presiding election officer is to fill in the results in Form 34A and electronically transmit the scanned Form 34A to the CTC using the KIEMS kits. The returning officer at the CTC is then to verify and tabulate the results from the various polling stations in the constituency and then, using these, generate form 34B. The results are then to be sent to the NTC, where the chairperson of the IEBC is to follow the same process of verification and tabulation in generating Form 34C. Physical Forms 34A are then delivered to the NTC by the returning officers for comparison with the scanned forms.

The audit, supervised by the Supreme Court Registrar, ICT experts and agents of the parties, confirmed various discrepancies.⁵⁵ These include the fact that a significant number of the electronically transmitted election results were not accompanied by scanned images of Form 34A, contrary to Section 39(IC) of the Elections Act.⁵⁶ This illegality was said to be compounded by various issues: the fact that the results were announced on the basis of Forms 34B before receiving all the necessary Forms 34A; the results on the IEBC portal differed from those on the Forms 34B provided to the Court; there were variations between results displayed on the IEBC portal and those in Forms 34A and 34B; and manipulation of the IEBC electronic system by third parties that generated numbers for transmission to the NTC.⁵⁷ Indeed, the IEBC admitted that it had declared the election results before receiving results from 11,883 polling stations, 17 constituency tallying centres and authentic Forms 34A from 5,015 polling stations, which represented more than 3.5 million voters.⁵⁸ It was, therefore, the petitioners’ contention that the transmission of unverified results flouted the principles of free and fair elections provided for in Article 81(e) of the Constitution as read together with various provisions of the Elections Act, the Elections (General) Regulations and Section 25 of the IEBC Act.⁵⁹

These discrepancies cast doubt on the integrity of the electoral process and raised questions about whether the election truly represented the free expression of the people’s will, as envisioned in Article 38 of the Constitution.⁶⁰ Consequently, the Supreme Court made an unprecedented decision on 1 September 2017 to nullify the presidential election. Despite the efforts to improve the electoral process through technology, these discrepancies and irregularities raised concerns about the credibility and integrity of the election results. These issues further contributed to the prevailing sense of voter apathy, as citizens became more sceptical about the electoral process and its ability to reflect the true will of the people. As noted by political scientist Nic Cheeseman, misinformation in relation to the 2022 general election results appeared credible due to a history of failure of technology in 2013 and 2017, as it played into public expectations.⁶¹

13.4 The Emergence of Disinformation and Misinformation: Impact on the 2022 General Election

In the lead-up to the 2022 general election, the IEBC worked diligently to address the challenges related to the use of technology in administering elections. But, at the same time, a concerning trend emerged, as politicians began adopting new strategies to influence political outcomes, capitalising on the rise of disinformation and misinformation. False narratives were strategically disseminated online by politicians with the intention of discrediting opposing candidates and gaining an advantage in the election. An example of this includes the spread of false claims, such as alleged defections or fabricated news of a candidate’s demise, often strategically timed on the day of the election or in the days leading up to it.⁶² This development aligns with Cheeseman’s and political scientist Brian Klaas’ observation that it has become increasingly difficult to detect election rigging. This is because whenever monitors devise new strategies to counter tried and true rigging tactics – in the case of Kenya, leveraging technology to prevent election rigging – dictators and despots continue to innovate.⁶³

13.4.1 Examining the Phenomenon of Disinformation and Misinformation in Kenya

Beata Martin-Rozumiłowicz and Rasťo Kužel proffer an apt definition of disinformation as deliberately ‘false or misleading information that is created or disseminated with the intent to cause harm or to benefit the perpetrator’ and misinformation as ‘false or misleading information that is shared without the intent to cause harm or realization that it is incorrect’.⁶⁴ In the context of elections, deliberately false information is spread to achieve political or financial gains and is usually directed towards either individuals, or groups, institutions and processes.⁶⁵ This section places its central emphasis on the utilisation of disinformation and misinformation, used interchangeably with the term fake news – defined as ‘purposefully crafted, sensational, emotionally charged, misleading or totally fabricated information that mimics the form of mainstream news’⁶⁶ – as a strategic tool wielded by politicians and various stakeholders to manipulate and shape the political narratives and results within Kenya’s electoral landscape. The discussion also covers instances where credible news is illegitimately labelled as ‘fake news’.

Misinformation is ‘one of the greatest challenges facing democracy in our time’.⁶⁷ In democratic societies, information is a powerful tool, especially during election campaigns. Kenyan elections exemplify this phenomenon, with a significant rise in digital campaigning across popular social media platforms such as Twitter and Facebook, messaging apps such as WhatsApp, and the video-sharing app, TikTok, during general elections.⁶⁸ At one end of the spectrum, this increases opportunities for the participation of women and youth, coordination by civil society and opposition, and undermines governmental control of information flow.⁶⁹ Social media has become an inseparable part of African politics, where politicians leverage influential bloggers to bypass mainstream media and directly reach voters with their campaign messages.⁷⁰ However, at the other end of the divide, this approach often involves collaborating with global firms, alongside local tech-savvy teams that engage in disinformation campaigns against election opponents.⁷¹

The potency of misinformation relies heavily on the extent of internet penetration within a country. As of 2022, reports indicated that internet penetration in Kenya stood at approximately 85 per cent according to Reuters, while Article 19, a human rights organisation, estimated it to be even higher at 93.7 per cent.⁷² In addition, there were around 11 million social media users, and this number continues to grow daily.⁷³ Unfortunately, Kenya’s low literacy rate means that internet users can be susceptible to believing and sharing fake news, turning ‘uninformed voters’ into ‘misinformed voters’.⁷⁴

The prevalence of big data in this digital landscape opens the door for foreign companies to engage in user profiling and targeted election campaigns online, exploiting the ease with which individuals can be tracked through social media platforms. The utilisation of information in elections is not a novel concept, and some argue it is akin to ‘traditional media advertising’, which employs surveys and focus groups to identify the most effective approaches for pitching newspaper and TV advertisements to voters.⁷⁵ The use of data in elections and political campaigns is also likened to the data practices used in marketing and commercial advertising.⁷⁶ Data companies often claim that the collection of information for profiling purposes is solely aimed at tailoring campaign messaging to suit individual voters, but this assurance raises apprehensions regarding the extent to which voters’ actions and beliefs are recorded and manipulated to achieve specific outcomes for particular candidates. This concern is particularly pronounced in authoritarian contexts, where the ruling elite benefit more from such practices due to their greater access to campaign resources and the government’s ability to restrict the entry of employees from companies supporting the opposition by denying them visas.⁷⁷

Modern autocrats increasingly find digital manipulation of data and information to be a useful tool, not only because it offers distinct advantages but also because it falls into a legal grey area, escaping strict regulation as election rigging.⁷⁸ This makes it an appealing option for those seeking to influence elections discreetly. As Charles Fombad aptly observes, this strategy makes ‘old techniques such as vote-buying, ballot-box-stuffing and double-voting seem positively crude and outdated’.⁷⁹

Mounting concerns surround the increasingly sophisticated methods data companies use to collaborate with candidates in engineering electoral victories through leveraging information to their advantage.⁸⁰ In Kenya, Cambridge Analytica, known for its data-driven approach to ‘changing audience behaviour’, was reported to have been involved in President Uhuru Kenyatta’s 2013 and 2017 general election campaigns.⁸¹ The company engages in mining user data from social platforms to target specific messages at individuals, based on profiles created from their internet usage, with the aim to encourage supporters to vote and discourage opponents from showing up to polling stations.⁸² These tactics provide a significant electoral advantage to a party, yet they are not explicitly prohibited by electoral law in Kenya, and they are not typically categorised as election rigging in election observation reports such as the IEBC’s Post-election Evaluation Report.⁸³ This raises concerns about the lack of regulation and oversight regarding such practices, and the potential impact on the democratic process and the integrity of elections.

There has been a significant growth in the complexity of disinformation campaigns, in terms of both the content disseminated and the channels used for distribution. Much misinformation is hard to prove or disprove as the information appears highly plausible to a broad range of people. Studies have shown that misinformation is more effective where it plays into existing prejudices or assumptions about a candidate, for example where it involves claims of corruption, incompetence, moral bankruptcy or failures as a spouse.⁸⁴

In Kenya’s 2022 general election, TikTok emerged as a major purveyor of false information.⁸⁵ Other studies reveal that Facebook was predominantly used for spreading campaign messages, while WhatsApp proved highly effective for mobilisation due to its extensive user base.⁸⁶ Aside from these, so-called political keyboard warriors emerged. These are bloggers who strategically position themselves during political campaigns to either create a favourable image of a candidate or de-campaign their opponents by spreading misinformation, often in exchange for payment.⁸⁷ Bloggers also spread misinformation to motivate candidates to hire them to rewrite the narrative and frame them in a positive light.⁸⁸

The combination of these tactics raises legitimate concerns about the manipulation of information and its impact on the democratic process, particularly considering that research indicates at least 75 per cent of Kenyan news consumers struggle to discern between real and fake news.⁸⁹ Further, a survey conducted in connection with the 2017 general election uncovered that approximately 90 per cent of respondents reported encountering content they suspected to be fake news, while around 80 per cent had come across information that they believed was intentionally fabricated or false.⁹⁰

Unlike other rigging tactics commonly used in hybrid democracies like Kenya, disinformation stands out as a strategy available to both the government and the opposition to manipulate electoral outcomes.⁹¹ Notably, there have been reports indicating that a group of Twitter influencers were allegedly paid by ‘shadowy financers’ in May and June 2021 to spread disinformation in favour of the government-backed Constitution of Kenya (Amendment) Bill 2020.⁹² This Bill was developed by the Building Bridges Initiative (BBI) Steering Committee and was famously held to be an unconstitutional attempt at constitutional amendment by the Kenyan Supreme Court, having been initiated by the President rather than being citizen-driven, which is what a popular initiative ought to be.⁹³ The disinformation campaign involved the harassment and discrediting of journalists, judges and civil society activists on Twitter, harming the individuals targeted and also undermining the credibility of the institutions they represent, which are important pillars of democracy.⁹⁴ The disinformation campaign had a clear objective: to promote the BBI process and falsely portray the aforementioned group of people (journalists, judges and so on) as villains opposed to development, and as being funded by William Ruto, the then Deputy President, who opposed the amendment. The aim was to deceive people, through hashtags and photoshopped images, into believing that the trending opinions were widely popular, and to create a false narrative surrounding the BBI process.⁹⁵

Commentators emphasise that the utilisation of data in elections and political campaigns is highly intrusive to individuals’ privacy and gives rise to significant security questions. Moreover, such practices have the potential to erode trust in the democratic process, as citizens may question the authenticity and legitimacy of election outcomes when data manipulation is at play.⁹⁶ As the use of data and digital tactics in elections continues to evolve, it is crucial to strike a balance between the legitimate use of voter information for campaign messaging and safeguarding citizens’ privacy and democratic values.

Concern has also been raised about the lack of accountability of global companies that engage in political campaigns within local contexts. A notable example of this occurred during Kenya’s 2017 general elections, where campaigns conducted by Harris Media LLC insinuated that presidential aspirant Raila Odinga would be involved in forcibly displacing ‘whole tribes from their homes’.⁹⁷ In a country with a history of electoral violence and deep ethnic divisions like Kenya, such online campaigns have the potential to stoke the embers of ethnic hatred and exacerbate social tensions.⁹⁸

13.4.2 Disinformation and Misinformation in the 2022 General Election

The 2022 general election witnessed the lowest voter turnout in fifteen years, with a mere 65.4 per cent of the registered 22.12 million voters participating, signalling an alarming surge in voter apathy.⁹⁹ The deployment of the KIEMS technology played a pivotal role in BVR and, during election day, in voter identification and results transmission. Notably, this system was successfully deployed and showcased improvements compared to prior elections.¹⁰⁰ However, the overarching technological concern during this electoral cycle revolved around the proliferation of disinformation and misinformation across online platforms.

During the election period, a deluge of misinformation plagued social media platforms, including false claims of victory, alleged political kidnappings, conspiracy theories and targeted attacks.¹⁰¹ A deeply concerning facet of the misinformation campaigns was the prevalence of sexualised gender attacks, particularly targeting female political aspirants, through platforms including Twitter, WhatsApp and TikTok. The disproportionately frequent attacks against women candidates further compounded the existing challenges they already grapple with during electoral campaigns.¹⁰² This situation was exacerbated by a critical factor: although the IEBC had posted images of Forms 34A illustrating results from approximately 99 per cent of the 46,663 polling stations across Kenya, the Commission had refrained from tabulating the cumulative figures on its public portal. The physical Forms 34A were being transported to the national tallying centre situated in the capital city, Nairobi, for comprehensive verification and tabulation of the grand totals before the official results were slated to be announced by the IEBC chairperson.¹⁰³

This informational gap engendered a confusing tallying process, as different media platforms broadcast conflicting figures based on their individual tabulation of Forms 34A. For example, on 10 August 2022, Citizen TV showcased presidential candidate Odinga leading with 51.3 per cent of the vote, with William Ruto in close pursuit at 47.3 per cent. Concurrently, the Nation Media Group displayed Ruto in the lead with 50.7 per cent, followed by Odinga at 48.6 per cent. This scenario sowed confusion and provided fertile soil for the proliferation of false news narratives. As recent studies show, the spread of misinformation is also problematic when it is sustained in and through traditional media, where traditional media fails to challenge misinformation.¹⁰⁴

A Mozilla Foundation study highlights that the spread of misinformation was further amplified by tech platforms, contributing to the uncertainty surrounding election results. Despite assurances from these tech companies that they would address problematic content in the run-up to the elections, the proliferation of misinformation remained a significant issue.¹⁰⁵ This failure to effectively combat the dissemination of false information on their platforms contributed to public confusion and distrust, as voters struggled to distinguish accurate information from misleading claims. On this issue, Cheeseman aptly observes:

Even those of us well seasoned in analysing rigging claims have struggled to cope with the bewildering monsoon of misinformation that has rained down on social media. When you wake up every morning to a new set of messages about how the process was manipulated, it can be hard not to believe there must be something in them – even as every claim you investigate turns out to be unconvincing.¹⁰⁶

After the announcement of Ruto as President-elect by then IEBC Chairperson Wafula Chebukati on 15 August 2022, with a margin of 50.49 per cent in the presidential vote compared to the runner-up Odinga’s 48.85 per cent, numerous parties contested the result of the presidential race. This situation became more pronounced following a last-minute announcement by four IEBC commissioners who disavowed the verification and tallying process, asserting that the actual results diverged from those about to be declared by the chairperson.¹⁰⁷ This dramatic turn of events occurred just prior to the chairperson’s planned announcement of the results, and was replayed on social media by prominent political leaders, causing a heightened state of agitation among Kenyans, and fuelling a surge of conspiracy theories online. As noted by a recent study, misinformation in relation to election results in 2022 appeared credible not only due to a history of failure of technology, which played into public expectations, but also due to its perpetuation by authority figures.¹⁰⁸ Where senior political leaders repeated what was not true, it gave false reports a second life by giving the misinformation political legitimacy.¹⁰⁹

The consolidated presidential election petitions were titled Raila Odinga and 16 Others v. William Ruto and 10 Others; Law Society of Kenya and 4 Others (amicus curiae).¹¹⁰ Similar to the preceding two post-2010 presidential election petitions, the contention arose that the deployment and utilisation of technology did not align with the prescribed constitutional and statutory standards.¹¹¹ In response, the IEBC asserted that the electronic election management system conformed to the constitutional criteria and that only authorised individuals could access the essential information, ensuring accuracy, completeness and protection against tampering, from both authorised and unauthorised sources.¹¹² Upon an order from the Supreme Court, the IEBC granted all petitioners supervised access to servers once the petition commenced. The Court also allowed for a thorough examination, recounting and scrutiny of the ballot boxes from the polling stations in question.¹¹³

The Court found that there was insufficient evidence to support the assertion that the technology deployed in the election administration failed to meet the prescribed standard set out in Article 86(a) of the Constitution, which encompasses transparency, security, integrity and verifiability.¹¹⁴ While acknowledging instances of KIEMS kit failures, the Court noted that manual verification successfully took place for the 235 polling stations where the KIEMS kits malfunctioned. It emphasised that Kenya’s election process is a blend of technology and manual methods, rendering the resort to manual voting for 86,889 voters compliant and not a disenfranchisement.¹¹⁵

It was held that the petitioners did not present any evidence that met the requisite standard of proof to demonstrate unauthorised access to the electronic election management system. The Registrar of the Supreme Court’s report likewise failed to indicate any security breaches in the Results Transmission System by unauthorised individuals.¹¹⁶ Addressing the petitioners’ claim of manipulation involving 11,000 Forms 34A before being uploaded onto the IEBC’s online public portal within eight minutes, the Court termed this assertion as practically implausible, as it would necessitate intricate coordination of people and alteration by individuals and machines across all 11,000 polling stations in question.¹¹⁷

In addition, the claim that certain Forms 34A presented to agents differed from those on the public portal was dismissed for lack of credible evidence. The Court disallowed the use of Forms 34A provided in affidavits by two advocates representing the first petitioner, which deviated significantly from the originals and certified copies on the public portal.¹¹⁸ The Court reprimanded the advocates for presenting misleading or fabricated evidence in the judicial process. The Court also noted instances of false and misleading information in the affidavit of prominent Kenyan corruption whistle-blower, John Githongo, including forgeries and inadmissible hearsays. Githongo’s application to withdraw the affidavit was denied by the Court.¹¹⁹ The Court’s frustration with the petitioners’ assertions, which ultimately held no evidential weight despite a rigorous trial process and thorough investigation, prompted the Court to label these claims as mere ‘hot air’ and liken their pursuit to a ‘wild goose chase’.¹²⁰ Concluding that the evidence was insufficient to warrant the annulment of the presidential election, the seven-judge Supreme Court bench unanimously dismissed the consolidated petition, thereby upholding the IEBC’s declaration of Ruto as the president-elect.¹²¹

As discussed, a concerning increase in the use of misinformation was observed after voting and during the hearing of the petition. It was argued that this increase was attributed to the fact that since voting was over, the potential for backlash from this misinformation affecting voting behaviour was reduced.¹²² Even more alarming was the alleged verification of the misinformation by trusted interlocutors such as Githongo, which made the misinformation more likely to be believed by Kenyans. However, little attention was paid to the potential impact of this misleading content on voter sentiment and behaviour, as well as the legitimacy of the new government and credibility of the IEBC.¹²³ The unchecked dissemination of misinformation can have serious consequences for the democratic process and the extent to which Kenyans can trust future electoral processes to be credible, warranting a closer examination of its implications on voter decision-making and the overall electoral landscape.

13.4.3 Impact of Disinformation and Misinformation during the 2022 General Election

With the rise in disinformation and unjustified labelling of credible news sources as ‘fake news’, there has been a concerning decline in public trust, affecting key institutions such as the traditional media (TV, radio and newspapers) and the IEBC. Kenya is already categorised as a low-trust society, with both political and social trust being at low levels.¹²⁴ Trust is a crucial element in the functioning of a democracy, and its absence poses a significant barrier to the consolidation of democratic values and principles.¹²⁵

The erosion of trust can be traced back to the aftermath of the 2007/2008 post-election violence and the subsequent disputed elections in the following electoral cycles, as earlier discussed. These events shook public confidence in the electoral process and raised doubts about the credibility of election outcomes. In particular, the nullification of the 2017 elections, which was overseen by the same chairperson as the 2022 elections, Wafula Chebukati, further contributed to the diminished trust in the IEBC. On this, it has been argued that:

Misinformation thrives, of course, when key institutions cannot be trusted and when it is repeated by respected figures. In the context of the Kenyan elections, many citizens went into the campaign with limited trust in Chebukati because he had presided over the 2017 elections that was nullified by the Supreme Court. Their trust was further eroded – some might say exploded – when, just as he was about to read out the result, four ‘rebel’ commissioners gave a rival ‘presser’ saying they could not stand behind the results. It did not matter that when the four commissioners subsequently provided details of their concerns it transpired that they rested, in part, on a mathematical misunderstanding so basic that it called into question both their capacity and their motivations. The sight of the IEBC once again at war with itself was enough to give credence to the claims the Commission had been politically captured and had fabricated the entire process.¹²⁶

The growing distrust of Kenya’s traditional media has created an environment conducive to disinformation and misinformation. Studies reveal a decline in public trust in mainstream media, which can be attributed to its perceived affiliation with ethnic and class interests.¹²⁷ When media outlets display discernible bias towards a particular candidate or political agenda, it further erodes public trust.¹²⁸ The concentration of media ownership in the hands of the political elite has strengthened the perception that the media is captured, prioritising the interests of political and corporate elites over its role in serving the public interest.¹²⁹ The media’s portrayal of varied election results following the 2022 elections only reinforced this perception of capture by political elites. The contradictory reporting by different media outlets contributed to the growing scepticism among the public, as they struggled to determine the accuracy and impartiality of the information presented.

Dubbed as ‘githeri media’ in a derogatory manner after its coverage of the 2017 elections, traditional media has lost its position as the primary agenda setter in the country, yielding ground to social media.¹³⁰ The term ‘githeri’ refers to a simple local dish made of maize and beans boiled or fried together. The label ‘githeri media’ is considered a symbol of disapproval, highlighting mainstream media’s fixation on trivial matters. This was evident during the 2017 general election when traditional media incessantly covered a man eating ‘githeri’ while overlooking more critical issues related to electoral irregularities and malpractices that were being discussed on social media platforms.¹³¹ This focus on the trivial over substantive electoral matters further eroded public trust in traditional media. However, the decline in public trust in media is an Africa-wide phenomenon.¹³²

With the public’s increasing scepticism towards traditional media, authoritarian figures and ruling elites find it easier to maintain their grip on power. They capitalise on discrediting independent media outlets that challenge the official state narrative. By undermining the credibility of those who question their authority, these elites can control the flow of information and shape the public perception in their favour.¹³³ Alongside the credibility deficit caused by the weakened trust in traditional media is the creation of a stratified society in terms of believability of information. As George Ogola correctly posits:

Significant too are the hierarchies that have emerged online. Those with the largest following tend to be politicians, celebrities, media personalities and those familiar with the digital literacy necessary to build a following. What these people say dominate conversations. The nature of agenda setting is thus personality driven. This is one of the most fundamental flaws of the online communication economy, and one that demands that we develop a deep circumspection.¹³⁴

Moreover, the rise of fake news and digitally manipulated information has given rise to what is known as a ‘post-truth’ world,¹³⁵ where objective facts are heavily contested, and voters may be swayed more by emotions than by verifiable information when making decisions on certain issues or leaders. In this environment, all information can be subject to doubt and scrutiny, allowing election-rigging claims to be disputed and enabling rigged elections to be falsely portrayed as legitimate.¹³⁶

In the past, authoritarian leaders had to exert significant effort to suppress information to maintain control. However, researchers now observe that many voters seem content with only receiving information provided by the government, effectively granting the government complete authority over defining truth and falsehood. This control over information empowers governments to manipulate facts at will, leading citizens to accept blatantly false statements.¹³⁷ As a result, the flow of information becomes so controlled by authoritarian regimes that they may no longer need to engage in overt rigging at the ballot box. Instead, they rely on indoctrination and fear tactics to achieve the same desired outcomes from the electorate.¹³⁸ In such a post-truth landscape, the erosion of public trust in reliable sources of information and the dissemination of disinformation create challenges for democratic processes. The acceptance of false narratives and the rejection of objective facts may hinder the ability of citizens to make informed choices, leading to the entrenchment of authoritarian rule and undermining the principles of free and fair elections.

Nevertheless, while fake news and misinformation remain challenges for democracies globally, misinformation did not reach the heightened levels many had feared in the 2022 Kenyan elections. Save for one doctored video of Ruto apparently attacking non-Kalenjin communities, which was quickly discredited by mainstream media,¹³⁹ the harder elements of misinformation such as use of deep fakes were absent. Kenya’s 2022 elections demonstrated that social media, through which a large portion of misinformation is spread, has not entirely changed the nature of elections in Kenya. Success in elections requires great investment in interpersonal campaigns, meaning that a candidate cannot rely exclusively on digital campaigns to sway the voting population. Nevertheless, it is unclear how much digital conversations affect behaviour offline or influence voting decisions.¹⁴⁰ Therefore, the intersection between digital media and pavement media in influencing opinions is an area that requires further investigation.

13.5 Detecting and Combating Disinformation and Misinformation: Legal Landscape and Engaging Stakeholders for Electoral Integrity

Given the detrimental impact of disinformation and misinformation on the foundations of democracy and the exercise of political rights, there is an urgent need to implement robust strategies for detecting and combating this pervasive problem. This urgency is underscored by Kenya’s historical context, marked by instances of election-related violence and manipulation, where the amplification of misinformation exacerbates the issues of voter disengagement, ethnic divisions and a general sense of bewilderment among the populace. In this section, we explore various recommendations that have been proposed, including legal, policy and institutional measures. Education and media literacy initiatives are also crucial to empowering citizens to discern accurate information from fake news and are discussed here. The section also highlights the significance of transparency, accountability and responsible use of technology to safeguard the legitimacy of elections and democratic processes. It thus emphasises the need to balance the use of technology, freedom of speech and media regulation.

13.5.1 Legal and Legislative Approaches to Combating Disinformation and Misinformation

In examining the legal and legislative measures to address fake news, it is essential to begin with Kenya’s Constitution. Article 33(1) of the 2010 Constitution guarantees the right to freedom of expression, including seeking, receiving and imparting information. However, this right is not absolute, as Article 33(2) clarifies that it does not protect expression that spreads propaganda for war, incites violence, constitutes hate speech, advocates ethnic incitement leading to harm or vilification of others, or promotes discrimination based on protected characteristics. In the context of Kenya’s elections, this would specifically involve combating the spread of ethnic hate speech, incitement to violence and defamatory statements that damage the reputation of others (Article 33(3)). Thus, a delicate balance is required between the right to freedom of expression, which includes freedom of speech, and protecting individuals from the harmful effects of misinformation and disinformation. In addition, Article 31 on the right to privacy also places limitations on freedom of expression by safeguarding individuals’ right to protect information about their families, private affairs and private communications.

Mainly applicable to journalists, Article 34(1) of the Constitution further guarantees the freedom of the media and the independence of electronic, print and other types of media. However, the limits set out in Articles 31, 33(2) and 33(3) also apply to the media. Article 34(3) allows for necessary regulation of the media to strike a balance between media freedom and preventing the dissemination of harmful information. The limits set in the aforementioned articles have, at times, been exploited to enable the unjust regulation of the media under the pretext of curbing unprotected forms of expression. Nonetheless, the courts have periodically struck down such provisions based on constitutional grounds.¹⁴¹

An additional constraint on freedom of expression and the media is encapsulated in Article 24 of the Bill of Rights, which constitutes the general limitation clause. Article 24(1) provides that fundamental rights ‘shall not be limited except by law, and then only to the extent that the limitation is reasonable and justifiable in an open and democratic society based on human dignity, equality, and freedom’. Subsequently, the latter part of the limitation clause (Articles 24(1)(a)–(e)) introduces a proportionality test, delineating five factors that must be considered to determine the appropriateness of limiting a particular right. These factors encompass ‘the nature of the fundamental right or freedom’, ‘the significance of the purpose of the limitation’, ‘the character and scope of the limitation’, ‘the necessity to ensure that an individual’s enjoyment of rights and fundamental freedoms does not impair the rights and freedoms of others’ and ‘the connection between the limitation and its purpose, along with the presence of less restrictive means to achieve the intended goal’. These constitutional provisions are reinforced by Article 27(4), which essentially prohibits misinformation and disinformation directly or indirectly discriminating against individuals based on various grounds, ‘including race, sex, pregnancy, marital status, health status, ethnic or social origin, colour, age, disability, religion, conscience, belief, culture, dress, language or birth’.

In addition to the constitutional measures, various statutes penalise the dissemination of certain kinds of fake news. One such is the Data Protection Act of 2019, which in section 2 defines data as any processed and recorded information by means of equipment operating automatically in response to instructions or as part of a relevant filing system. Section 25 of the Act sets out the principles and obligations of data protection, emphasising the respect for the right to privacy of the data subject and that personal data should be ‘collected for explicit, specified and legitimate purposes and not further processed in a manner incompatible with those purposes’.¹⁴² The Act also requires data controllers or processors to ensure that personal data is accurate and kept up-to-date, with measures taken to promptly erase or rectify any false or misleading data.¹⁴³ These provisions serve to protect against fake news that violates individuals’ privacy through the dissemination of personal data and the use of altered photos and videos, commonly known as deepfakes, to spread misinformation and disinformation.¹⁴⁴ As noted above, however, the use of deepfakes in the 2022 general elections was limited.

Section 26 of the Act further provides for the rights of erasure, correction or deletion of false or misleading data, as well as the right to be informed and consent to the collection and use of personal data. Meanwhile, section 28 permits the collection of data from public records when the same has been made public or with the data subject’s consent from another source. However, sections 28 and 29 also mandate that reasonable steps must be taken to notify the data subject that their data is being processed, and failure to do so is considered an offence. These provisions aim to address the spread of fake news that involves the manipulation of personal data and provide individuals with the right to manage their data, ensuring accuracy and protection from misuse.

Second, the Computer Misuse and Cybercrimes Act of 2018 addresses the intentional publication of falsehoods and misinformation. In section 22(1) of the Act, the deliberate dissemination of false information is criminalised, and those found guilty can face a fine of 5 million shillings (Ksh; $50,000) or a maximum of two years’ imprisonment, or both. Furthermore, section 22(2) of the Act limits freedom of expression, in accordance with Articles 24 and 33 of the Constitution, with regard to intentional publication of falsehoods and misinformation. The Act specifically targets content that is likely to incite violence, promote war, propagate ethnic hate speech or discrimination, and negatively affect the rights or reputations of others. Section 23 further addresses the intentional publication of false information with the specific intention of causing panic and chaos, or inciting violence. Those found guilty of spreading such false information can face severe penalties, including a fine of Ksh 5 million or a maximum of ten years’ imprisonment, or both. These legal measures serve to deter the spread of fake news and misinformation in the digital realm, ensuring accountability for those who intentionally disseminate harmful false information.

The constitutionality of the Computer Misuse and Cybercrimes Act was contested in the case of Bloggers Association of Kenya (BAKE) v. Attorney General and 3 Others; Article 19 East Africa and Another (Interested Parties).¹⁴⁵ With regard to sections 22 and 23 of the Act, which criminalise the dissemination of false information, the petitioners argued that truth was not a prerequisite for free speech. They further contended that these laws, which aimed to regulate truth using terms such as ‘false, misleading or fictitious data or information’, could infringe upon the freedom of expression.¹⁴⁶ In response, the Court reasoned that ‘the state has a legitimate interest in ensuring the safety and integrity of information and the protection of its citizens against cybercrimes’, especially when the freedom of expression impacts not just individual rights but also the broader public.¹⁴⁷

Considering the rapid dissemination of misinformation or fake news on the Internet, leading to fear, panic and even potential threats to national security, the limitations imposed by sections 22 and 23 of the Act in the form of the offence of publishing false information over a computer system were found to be reasonable, justifiable and in line with the constraints set out in Articles 24, 31, 33(2) and 33(3) of the Constitution.¹⁴⁸ These provisions permit publishers to express themselves as long as their expression does not harm others and avoids constituting cyber libel.

Addressing the claim that the terms ‘false, misleading or fictitious data or information’ are vague, overbroad and would have a chilling effect on speech by, inter alia, journalists, activists, academics and politicians, Justice Makau held that the language used in sections 22 and 23 is precise and clear when interpreted contextually according to statutory interpretation principles.¹⁴⁹ He further noted that the term ‘false’ is readily understood in plain English, and does not necessitate a legal definition. Its application in section 23 was said to be evident – to criminalise the dissemination of false information with the intention to induce panic, chaos, violence or tarnish an individual’s reputation.¹⁵⁰ As a result, the Court upheld the constitutionality of these two provisions.¹⁵¹ However, civil society actors continue to call for the repeal of sections 22 and 23 of the Act. They argue that ‘The vague prohibition of “false”, “misleading” and “fictitious” data is highly subjective, and this law has been used to harass journalists, bloggers and activists in the past’.¹⁵² They thus recommend the removal and replacement of these sections with less intrusive measures to combat disinformation.¹⁵³

Third, the Defamation Act offers civil remedies in cases where fake news qualifies as libel, as outlined in Section 7(3).¹⁵⁴ It is also important to note that Section 7(3) specifies that the Act does not protect the publication of any matter that is not of public concern and that does not serve the public benefit. The Act ensures that the law supports the freedom of expression regarding matters of public interest, while also holding individuals accountable for spreading false information that can harm others or undermine public trust in the media.

Fourth, Section 66(1) of the Penal Code criminalises the publication of false statements, rumours or reports that could potentially induce fear, alarm or disrupt public peace.¹⁵⁵ This section designates such actions as misdemeanours and assigned guilt to those perpetrating them. Nonetheless, if the accused person can demonstrate that they took reasonable measures to verify the accuracy of the news before publication, believing it to be true, Section 66(2) provides them with a legal defence.

In the case of Cyprian Andama v. Director of Public Prosecutions & 2 Others; Article 19 East Africa (Interested Party), the petitioner challenged the constitutionality of section 66(1).¹⁵⁶ The petitioner argued that the wording of this section, which states, ‘likely to cause fear and alarm to the public or to disturb public peace’, was overly vague and its coverage too broad, thus allowing for subjective interpretations and potential misuse in charging individuals. This, the petitioner contended, violated their Article 33 right to freedom of expression.¹⁵⁷ Furthermore, it was asserted that the provision was so unclear and uncertain that individuals could not discern the boundaries of permissible communication, potentially ensnaring both innocent individuals and those who were not.¹⁵⁸ The central issue before the High Court revolved around whether the limitations on freedom of expression outlined in Section 66 of the Penal Code could be deemed reasonable and justifiable in a free and democratic society, as mandated by Article 24 of the Constitution, which sets conditions for limiting fundamental rights and freedoms.¹⁵⁹

Justice Korir’s judgment established that Section 66 allowed for the conviction of individuals merely for making statements that were considered untrue, without placing the burden on the prosecutor to demonstrate the statement’s falsehood. This, in turn, created an atmosphere of trepidation surrounding the exercise of freedom of expression.¹⁶⁰ Moreover, the section was deemed excessively broad as it could potentially prohibit not only the publication of false statements but also opinions that were honestly believed to be truthful, violating the rule that legal provisions establishing criminal offences are to be clear, concise and unambiguous.¹⁶¹ The Court also noted that the Defamation Act offered a less restrictive means of safeguarding the reputation of individuals who had been unfairly defamed. It provided a civil remedy to restrain those intending to damage others’ reputations by inflicting monetary penalties through the award of damages.¹⁶² Consequently, the Court ruled that Section 66 was an unjustified violation of Article 33 and, therefore, unconstitutional.¹⁶³ A cautious approach is thus necessary, as the aforementioned laws, which ostensibly limit freedom of expression and speech, have sometimes been exploited by the government to suppress critics and impede the open discourse of political matters, especially in the context of elections.¹⁶⁴

While Kenya leads the region in the regulation of data protection and abuse of social media platforms, it is apparent that electoral law, as currently formulated, does not regulate the increased use of social media in elections. While election laws prohibit campaigning two days before a general election, there is no law that bans candidates from purchasing advertising on social media platforms up until election day.¹⁶⁵ Also of concern, online advertising is not adequately regulated as the Electoral Code of Conduct does not cover online activities. Since the IEBC has proposed to undertake a review of its Electoral Code of Conduct following several court challenges arising from the 2022 elections,¹⁶⁶ we propose that the IEBC engage stakeholders in the development of specific regulations for online campaigns, with attendant sanctions also included in the Election Offences Act.¹⁶⁷ The Kofi Annan Foundation similarly recommends signing of the Digital Pledge by all political actors to stop the spread of misinformation.¹⁶⁸

13.5.2 Institutional Measures to Combat Disinformation and Misinformation

Institutional measures are equally important in tackling disinformation and misinformation. Social media sites and tech companies have an obligation to moderate content to reduce misinformation and disinformation. However, platforms such as Facebook have been criticised on the basis that the allocation of resources towards misinformation (estimated at 87 per cent in 2021) is disproportionately allocated to English-language content, yet only 9 per cent of its users actually speak English.¹⁶⁹ Odanga Madung attributes this to an unwillingness by tech platforms to ‘devote time and resources to an electoral process outside their cultural context’.¹⁷⁰

Furthermore, Article 19 has expressed concern over the lack of feedback incorporation from local communities in the content moderation strategies of social media companies. As a result, these strategies often fail to consider the cultural, social, historical, economic and political contexts of the communities they affect.¹⁷¹ Most language moderation tools primarily cater to English, rendering them ineffective in instances where local slang or a mixture of multiple languages like Kiswahili and English are used. Further, these tools struggle to interpret statements involving derogatory metaphors or those heavily dependent on contextual understanding.¹⁷² The absence of comprehensive content moderation facilitates the spread of misinformation and exacerbates its impact on electoral processes.

In 2022, the IEBC entered into memorandums of understanding with the Media Council of Kenya and tech companies, such as Facebook, on regulating online content. It may be necessary to create a coalition of actors to address concerns in this area. Having successfully convened political actors under the Political Parties Liaison Committee, IEBC can use existing goodwill to convene social media platforms to address digital threats to electoral integrity. As proposed by the 2020 study by the Kofi Annan Foundation, having such a coalition would allow for the creation of ‘cross-platform strategies for detecting and limiting the reach of weaponized disinformation and hate speech’.¹⁷³ Collaboration between the IEBC, online platforms, such as Meta and Twitter, along with local information fact-checking organisations, presents an effective strategy to combat disinformation and misinformation using both reactive and proactive measures. This can involve actions like suspending accounts linked to disinformation campaigns. In addition, implementing measures such as Twitter’s suspension of the trending feature during elections, as observed in Ethiopia during a period of rising disinformation and hate speech, can contribute significantly to countering the issue.¹⁷⁴

13.5.3 Non-Legal Initiatives to Combat Misinformation and Disinformation

Several initiatives have been undertaken to address the question of fake news. Africa Check is a civil society initiative that works across Africa to verify news content and political statements and is part of the International Fact Checking Network (IFCN) which brings together fact-checkers and advocates for factual information, working globally to combat fake news.¹⁷⁵ However, these sites get less visibility than those that spread false information. This limits their reach in addressing disinformation and misinformation.¹⁷⁶ Moreover, a ‘tidal wave of misinformation’ makes it difficult to keep up with fact-checking and challenge fake news.¹⁷⁷ Therefore, a collaborative approach aimed at enhancing the capacity of organisations already engaged in real-time information fact-checking, particularly in the context of elections and crucial political matters, becomes essential.¹⁷⁸

Further, the use of artificial intelligence (AI) tools to track misinformation in relation to elections has been demonstrated to be ineffective due to the language settings of these tools as well as their inability to identify context-specific misleading information. Existing tools have been criticised for not being tailor-made to local contexts. It is recommended to find ways of refining AI tools to make them better able to address local languages, in collaboration with local actors. However, since it is difficult to recall misinformation once it achieves virality, it is recommended to increase human review for accounts or posts that threaten the integrity of elections.¹⁷⁹

Education and media literacy initiatives are also crucial to empowering citizens to distinguish accurate information from misinformation, thereby addressing the human vulnerability that enables the propagation of fake news. As an example, the cooperative effort among stakeholders to provide educational initiatives aimed at fostering critical thinking in the digital realm can be exemplified by Meta’s ‘trusted partner’ programme. This programme serves to assist civil society in mitigating online and offline risks by offering expert insights and recommendations.¹⁸⁰ Increasing critical thinking and media literacy can make individuals more resilient to the influence of false information and less susceptible to manipulation. Importantly, it is reported that voters who spend more time on social media platforms tend to have a higher capacity for fact-checking and display a greater level of scepticism.¹⁸¹ This heightened awareness makes them less susceptible to disinformation and misinformation campaigns.

In addition, debunking of misinformation related to the 2022 elections also limited its spread on digital platforms and vastly reduced its effect. Even as more work is done in retraining AI models to make them suitable for local contexts, where political actors, traditional media and other influential personalities call out misinformation frequently, it increases the level of scepticism among the voting population, making it harder for them to be deceived.¹⁸²

13.5.4 The Need for a Balancing Exercise

A key aspect to consider is the balance between the effective use of technology in elections and the regulation of abusive behaviour. As this chapter has discussed, technology has the potential to enhance the integrity and credibility of elections, but it can also be exploited to spread disinformation and manipulate public opinion. Striking a balance between leveraging technology for democratic advancement and curbing its misuse is essential. Furthermore, there is a delicate balance to be struck between upholding privacy and security rights and protecting freedom of expression and political rights. Any regulatory measures should carefully navigate these contrasting rights to avoid unnecessary restrictions on freedom of speech while safeguarding against the misuse of data for disinformation and misinformation campaigns.

13.6 Conclusion

As noted, technology can serve as a tool to enhance democratic legitimacy and combat election rigging when properly implemented and managed. While social media is now an integral part of our elections, and cannot be disregarded, the dark side of the use of technology to spread disinformation and misinformation remains a threat to democratic trust and legitimacy, and its effects are yet to be fully appreciated. To restore public trust in key institutions protecting democracy, and in the integrity and legitimacy of elections, a comprehensive and multifaceted approach is necessary to combat disinformation and misinformation. Legal, policy, institutional and educational measures should work in harmony to preserve the integrity of elections, and protect democratic values. Striking the right balance will be key to fostering an informed and engaged citizenry, ultimately strengthening democratic processes and rights, and preserving the democratic values of open debate, informed decision-making and the safeguarding of public trust in the democratic process.