An analysis of threats to Kenya’s media landscape ahead of the August 2027 elections finds that AI deepfakes, synthetic media, and coordinated online campaigns threaten voter choice and democratic integrity. Across Africa, emerging regulation, verification tools, and fact-checking efforts are constrained by limited funding, weak platform accountability, and largely reactive debunking.
Ahead of the August 2027 Kenyan elections, media landscape stakeholders worry that the rise of fake news and information disorder, amplified by Artificial Intelligence (AI), could undermine citizens’ right to reliable information needed to make political choices.
This also happens in other countries across the world, amid weak online platforms’ accountability that facilitates information disorder – misinformation, disinformation and malinformation.
Information disorder rises as news consumption shifts to digital media and the struggling local press faces financial challenges. Moreover, perceived political influence leads to substandard journalism. Malicious actors seize the opportunity to spread misinformation and disinformation online. This has happened before in Kenya, Africa, and other parts of the world.
The misuse of social media platforms, gaps in platform accountability, and limited fact-checking have threatened information integrity and fueled the rise of fake news. This often happens during electoral periods and political crises – protests, coups, and other forms of political instability. AI-generated content has complicated access to reliable information online. Some paid influencers, unethical journalists, and content creators are spreading false information through organised campaigns to promote misleading narratives.
The high volume and sophistication of AI-driven, targeted disinformation could undermine citizens' ability to form their own political opinions and make informed decisions at the ballot box, warns a recent research article published by the Konrad Adenauer Stiftung.
Titled "The Rise of Fake News Ahead of the 2027 Kenyan Elections", it documents “sophisticated” and escalating information disorder. This threatens democracy and citizens’ ability to make informed decisions. In Kenya and other parts of the world, news and information consumption has shifted towards digital and social media platforms like TikTok and WhatsApp. This shift has bypassed gatekeepers and created a complex environment where activism and organised disinformation overlap, according to a recent article by the Konrad Adenauer Stiftung.
AI-generated content has complicated access to reliable information online. Some paid influencers, unethical journalists, and content creators are spreading false information through organised campaigns to promote misleading narratives.
As a result of information integrity challenges, hate speech, misinformation, and disinformation dominate political discourses, thereby undermining elections. Information disorder also limits access to credible, factual and pluralistic information about candidates, political parties and issues. For instance, 2024 was dubbed the year of Democracy in Africa – dozens of countries went to the polls. However, several governments deployed digital tools to compromise information integrity, spreading their narratives to stay in power.
AI tools were used in Rwanda and South Africa to create deepfakes and synthetic media, manipulating public perception. Critics accuse social media platforms of being slow to address this issue, and where they have, the approaches have not been uniform across countries, notes the Collaboration on International ICT Policy in Eastern and Southern Africa (CIPESA) in its 2024 State of Internet Freedom report that focused on the rising digital authoritarianism and the use of technology to win unfair elections.
Now, the Media Council of Kenya, the country’s national media regulator, has included AI in its recently unveiled 2027 media coverage guidelines for general elections, signalling the crucial role AI plays in information integrity. For instance, it can be used to manipulate information for various purposes. The guidelines call for responsible AI use, stating it should not edit electoral information.
AI-Enabled Disinformation
Limited digital and information literacy has facilitated the spread of AI-generated content used for disinformation and misinformation. Such content undermines access to reliable information and public perception. This is accelerated by limited skills in fact-checking and verification of misleading content. This includes videos and images used out of context to manipulate and spread disinformation. Other content is created using AI tools to mislead the targeted audience.
AI enables the rapid production of sophisticated disinformation that many actors in Africa and beyond may be unable to identify. Ahead of the 2024 election year, CIPESA's annual report noted concerns about how AI could harm the polls. It gives examples such as using generative AI to impersonate election officials, cloning election results management processes, and undermining citizens’ privacy rights through access to voters' data.
The use of AI in election disinformation was cited in Rwanda and South Africa. “Synthetic media, including deepfakes, were evident in many elections, taking some new forms from previous polls – such as allowing easy translation of information, swift generation of fake videos, manipulation of images, and arming of bots.
"CIPESA Reveals. Synthetic media refers to content that is artificially created or manipulated using digital technologies, often leveraging AI.
AI enables the rapid production of sophisticated disinformation that many actors in Africa and beyond may be unable to identify. Ahead of the 2024 election year, CIPESA's annual report noted concerns about how AI could harm the polls. It gives examples such as using generative AI to impersonate election officials, cloning election results management processes, and undermining citizens’ privacy rights through access to voters' data.
Elections and Information Integrity
In the past, Kenya has experienced disinformation during elections in what experts term a professionalised “disinformation industry”. The weaponisation of information not only deepened ethnic and class divisions but also contributed to profound voter apathy, resulting in the 2022 elections recording the lowest voter turnout since 2002, notes Anna Lisa Hildebrandt in a recent research article on the rise of fake news before and during elections.
Concerns are now increasing ahead of the 2027 elections over the rise of fake news and online propaganda. The proliferation of unregulated social media platforms has complicated the matter. Hate speech and algorithmic amplification of political propaganda pose a major risk to information integrity. Media landscape players will have to double their efforts in combating information disorder.
Consistent fact-checking and digital tools that enable information verification will be crucial to debunking claims and fact-checking news and information, protecting information integrity. Other countries worldwide are doing the same. Information disorder is now a global challenge.
Reactive Fact-Checking
Given the robust disinformation in Kenya, across Africa, and around the world, legal, technical, and fact-checking tools will play a key role in combating information disorder. However, the challenges remain. For instance, in some countries, “The legal landscape for combating disinformation in Africa is increasingly shaped by a securitisation of information, where false narratives are treated as direct threats to national security."
"Warns Konrad Adenauer Stiftung.
In recent years, fact-checking organisations have risen across Africa and around the world. Those initiatives have been combating information disorder by debunking viral claims. However, those fact-checkers face several obstacles, such as funding gaps and sometimes limited collaboration with global social media platforms. In Africa, some of the leading fact-checking initiatives include Africa Check and PesaCheck.
Another major challenge is reactive fact-checking. Research indicates that fact-checking in Kenya is primarily reactive, meaning that by the time a claim is debunked, the original disinformation has already reached its peak audience.
Some local media organisations are making significant efforts to fact-check regularly. For instance, Radio 47 is a local commercial station in Kenya that broadcasts in Kiswahili, a language spoken by most Kenyans and by followers abroad. The station fact-checks online content – videos, images and text, mostly from social media. It debunks the content after its news programmes. This practice can be emulated by other local and foreign media outlets.
Some international media organisations have also set up fact-checking desks. One such initiative is the AFP fact-checking desk. However, they face challenges, including limited funds and limited collaboration with social media platforms. For instance, on August 14, 2024, Meta ended access to the decade-old ‘CrowdTangle’, a popular public insights tool for exploring social media content and tracking misinformation. It replaced it with the Meta Content Library and API, which some stakeholders claim is “less transparent, restricted to academics and non-profit researchers, accessing it is cumbersome and slow, and it has fewer features than CrowdTangle.”
At the time, Sophie Nicholson, Deputy Chief Editor at Agence France-Presse, Fact-Check Desk, was quoted as saying, "It is very concerning, because there’s nothing else out there right now that we can see that would be as useful or as efficient in our fact-checking work.” This was in an article on Facebook's decision to close CrowdTangle.
Use Verification Tools
For effective election information monitoring, various tools have been suggested by fact-checking and media professionals. Such tools allow users to fact-check viral claims and other forms of disinformation.
One such tool used in several countries is 'I verify'; the platform supports actors worldwide in preventing and mitigating disinformation, misinformation, and hate speech. Managed by the United Nations Development Programme, national counterparts implement it to verify information flagged by the general public. The tool is being used in the Democratic Republic of Congo, Honduras, Kenya, Liberia, Madagascar, Nigeria, Pakistan, Sierra Leone, and Zambia.
Other tools are being created; for instance, Delphine Nyaboke, a Mozilla fellow, is developing Bias Lens, a tool designed to flag bias before a story or an AI-assisted draft goes live. The developer says it will use AI to detect misinformation, misleading framing, cultural or gender bias, or missing context altogether in Kenyan elections, and that it could potentially be deployed in other African countries.
Some measures implemented in South Africa to combat hate speech, misinformation, and disinformation ahead of the elections include the Real411 election reporting portal and various fact-checking initiatives by the media and independent fact-checkers such as Africa Check and AFP.
Initiatives to Address Disinformation
Cipesa's 2024 report cites several initiatives across Africa to protect information integrity, especially those used during the 2024 elections.
Some measures implemented in South Africa to combat hate speech, misinformation, and disinformation ahead of the elections include the Real411 election reporting portal and various fact-checking initiatives by the media and independent fact-checkers such as Africa Check and AFP.
Similarly, in Rwanda, ahead of the July 2024 elections, a coalition of local media organisations and Africa Check partnered to fact-check information, ensuring voters received accurate information to make informed decisions. In Chad, Meta held a workshop in April 2023 to raise awareness of its community standards on disinformation management.
To sum up, the fight to protect news and information integrity, especially during elections, will involve strong and independent media organisations, constant fact-checking, advanced media and digital literacy for citizens, stable and strong fact-checking organisations, platform accountability to address context bias and increase content moderators from the Global South, and increased use of fact-checking tools to debunk online claims.
