Social media illustrationKenya’s political information space is undergoing a major shift as influencers, algorithms, and emerging artificial intelligence tools increasingly shape how voters understand politics.
A new study by the Kenya ICT Action Network (KICTANet) says alternative media's growing influence is outpacing the systems designed to hold political communicators accountable.
As such, there are concerns over polarisation and declining trust in institutions ahead of the 2027 general election.
The report dubbed Many Microphones, Little Accountability, says the problem has escalated beyond the circulation of false information to a political communication system where personalities command large audiences without the editorial checks and balances traditionally associated with journalism.
The findings are based on monitoring of Facebook, TikTok and YouTube between January and March 2026 focussing on political information consumption and verification. The study involved interviews with journalists, content creators, civil society and policy actors.
Although researchers caution the study is not statistically representative of the entire digital ecosystem in Kenya, its findings point to a significant change in where political influence is being exercised.
The report says trust is increasingly shifting from institutions to personalities. Influencers are often viewed as more relatable, accessible and authentic, even when they may not be more accurate.
“People believe the messenger they relate to, not the institution behind the message,” one key informant told the researchers.
Influencers are now shaping how audiences interpret political events before they encounter verified information from electoral commission, government agencies or established news outlets.
The 2026 Reuters Institute Digital News Report similarly found that Kenya stands out for the prominence of creators in the news ecosystem. From the research, 33 per cent of respondents said creators meet all or most of their news needs. TikTok and YouTube have become important entry points to news, particularly among younger audiences.
KICTANet found the influence is reinforced by the way platforms reward content that attracts strong emotional reactions.
Anger, humour, fear and political loyalty can generate engagement, giving emotionally charged political narratives greater visibility than slower, evidence-based reporting.
On TikTok, researchers found that selective video clips, captions and commentary could compress complicated political issues into simple conclusions. On Facebook, fabricated newspaper front pages and fake articles borrowed the authority of established media brands.
One case documented by the report involved a fabricated Standard article about divisions within the opposition. PesaCheck established that the article had never appeared on The Standard’s website or its official social media accounts. This is also common among other mainstream publications.
The report categorises the process of development of such information into three phases: fabrication, acceleration and consolidation.
A fabricated claim can emerge on Facebook, gain speed through TikTok and later be discussed at length on YouTube or podcasts, giving an unverified allegation the appearance of an established political narrative.
The economic incentives of digital media can further reinforce the cycle. KICTANet says content creators face incentives to maximise views, clicks and engagement, with one interviewee noting, “controversy drives engagement, and engagement translates to income.”
The report warns that repeated exposure to politically aligned content can deepen echo chambers and polarisation, while declining confidence in institutional sources can make official corrections less effective.
KICTANet says corrections and clarifications frequently reach smaller audiences than the original fabricated material. It describes the institutional response as reactive, fragmented and poorly coordinated, with particular weaknesses in responding to vernacular and county-level content.
This finding corroborates past research on the 2022 election, which found coordinated influencer campaigns, manipulated images, fabricated quotations and false election-related narratives circulating across social media.
A recent academic study described a structural “timing trap” in which false information travels faster than professional fact-checking can respond.
The next election, however, could add another layer due to AI-generated content.
KICTANet warns that deepfakes and synthetic media could make political manipulation more convincing and harder to detect.
A 2026 survey experiment involving 7,000 Kenyans also found that social-media commentary can affect how audiences interpret the authenticity of political deepfakes.
KICTANet recommends establishing a multi-stakeholder election information integrity task force involving the IEBC, Communications Authority, Media Council, technology companies, civil society and fact-checkers. It also called for stronger local-language moderation, faster responses to fabricated content and nationwide digital and media literacy campaigns.
