The role of AI-Curated Content on Anti-Government Sentiments Among Gen Zs in Kenya
Description
This study examines how algorithmically curated social media feeds influence political perceptions and activism of Kenyan Gen Zs, a generation that leads online political activism in Kenya. This study is important in that it places the Kenyan and African youth in global debates surrounding the influence of algorithmic social media in promoting political dissent and democracy, discussions which tend to ignore African contexts. Based on agenda setting theory and social identity theory, algorithmic feeds are treated as active agenda-setters, promoting issue prominence, while at the same time reaffirming group identities. A quantitative survey methodology was used with data obtained from 615 respondents spread across all eight regions in Kenya. Logistic and ordinal logistic regressions were utilized for data analysis. Exposure to negative content was found to be a strong predictor of belief in change, concern, and political protest while personalization of the algorithm reduced these effects. The study recommends making the algorithms more transparent, initiating digital literacy campaigns, carrying out impact assessments on social media platforms, and encouraging multi-stakeholder management.
Files
Steps to reproduce
The study population consisted of Kenyan Generation Z members, defined in this case as members of Generation Z from Kenya who were between the ages of 18 to 26 years in the year 2025 and who use one of the top three social media platforms (Facebook, TikTok and YouTube). The sampling method used was stratified random sampling. The strata were based on the eight regions that make up Kenya historically (namely, Nairobi, Central, Rift Valley, Western, Nyanza, Coast, Eastern, and North Eastern regions). These eight regions were used because they represented diversity socially, culturally, economically, and politically since each region maintained unique media and political pasts. Primary data collection was done using a structured survey that was conducted both online and physically. Below are how the key constructs were operationally defined. Exposure to Negative Content: Defined using a five-point Likert Scale (Never to Very Often) to measure the question: “How often do you come across posts on [platform] criticizing the Kenyan Government or talking about serious issues facing Kenya?” Perceived Algorithmic Personalization: Defined using a five-point scale measured by the question: “How often do you think that the content in your social media feed is personalized based on your prior activities or interest?” Systemic Change Belief: Defined using a Yes/No binary outcome measured by asking: “Do you think Kenya requires a total systemic change politically and/or in its form of governance?” Social Media Fuelled Protest: Defined using a Yes/No binary outcome measured by asking: “Have you ever been at a protest or public demonstration whose awareness came through social media?”