Towards Automated Fact-Checking in Africa: The Experience With Artificial Intelligence at Africa Check

Abstract This research examines the evolution of automated fact-checking at Africa Check, a prominent fact-checking organisation based in South Africa, Kenya, Nigeria and Senegal, through the integration of artificial intelligence (AI) tools into its fact-checking process. The progress focuses on three phases, which include claim detection to identify check-worthy claims, claim matching to retrieve already verified information and robot-checking to validate claims. The methodology used in this study involved analysing the experiences of reporters and editors from Africa Check. In-depth interviews were conducted to reveal the advances in time and effort required to verify information, providing key insights into the relationship between technology and journalism. The findings also shed light on how AI can be used to combat misinformation and promote media literacy in Africa, demonstrating how the adoption of AI has helped Africa Check to expand its reach and increase its impact. As a result, these improvements have profound implications for political accountability, trust in the media, and the fight against misinformation. This research offers valuable contributions to the ongoing debates about the adoption of technology in journalism and the importance of leveraging technology to strengthen the relationship between journalists and their audiences.