AI, Journalism, and Public Interest Media in Africa

"The integration of AI into newsroom practices and processes will depend on how well AI is understood by journalists, editors, and media managers..."



This study, published by International Media Support (IMS), highlights the state of artificial intelligence (AI) use in Africa's media. It explores AI's potential in strengthening public interest media through selected country case studies from Eastern, Southern, and West Africa. It also highlights some of the challenges and offers recommendations for the way forward. This is the second study to scope AI use by media in regions of the world where IMS operates. The first study looked at the use of AI and machine learning in Latin America and Eastern Europe (see Related Summaries, below).



The research sought to answer the following questions:

• How well is AI understood within the African media landscape?
• How are AI technologies used in Africa's public interest media? Are they changing or reconfiguring everyday newsroom practices and routines across the continent?
• What are some of the challenges facing AI adoption in Africa's public interest media, and how can these be addressed?
• What is the state of data accountability mechanisms including legislation, policy, and governance in the continent?
• What are the prospects for AI use in Africa's public interest media?

The study used in-depth qualitative key informant interviews (KIIs) of key stakeholders/informants, newsroom observation, and document analysis. The key informants included news reporters, news editors, digital editors, audience engagement editors, media managers, data specialists, fact-checkers, and media scholars. In Eastern Africa, the research looked at newsrooms and other media actors in Kenya, Tanzania, Uganda, and Somalia. In Southern Africa, it looked at Zimbabwe, South Africa, Zambia, and Namibia, while in West Africa it focused on Nigeria and Ghana. Researchers used purposive sampling primarily because they wanted to focus in depth on a limited number of samples.



The study found that while there is AI use across several news organisations in Africa, the adoption of AI systems and tools in African media remains relatively low. AI use also varies regionally and across different media, with the most use cases in Kenya and South Africa. The big well-resourced media have invested in several premium AI systems and are also developing several custom-built AI tools. However, most of the smaller media organisations have either not adopted AI into their newsroom processes, or where they have done so, they rely largely on open-source tools. The most used AI systems are functional AI, deployed in content/news gathering, content processing, content/news distribution and audience engagement, and various editorial practices.



Key challenges facing AI adoption include a notable gap in knowledge about AI, resource constraints, fear of the negative impact of algorithmic systems and professional insecurity among journalists, cultural resistance within newsrooms, marginalisation of women in the newsroom, weak policy and legal infrastructure, lack of appropriate AI business strategies, the existence of "dirty" data, and a lack of collaboration among news actors in the continent.



The study concludes that while technology is not the panacea for the challenges facing African journalism, AI holds significant potential to strengthen the production and distribution of public interest journalism throughout the continent. It offers several recommendations, which focus mainly on how international media development organisations can support African media in harnessing the potential of AI. These include:

• Supporting capacity-building initiatives to address the knowledge gaps - This could include working with journalism schools to develop curricula that address AI literacy in the newsroom.
• Tackling resource challenges - For example, media organisations should be encouraged and assisted in developing business cases for AI in their newsrooms.
• Enhancing collaboration - For example, media organisations should be encouraged to collaborate with one another, media/digital innovation hubs, universities, and the wider tech community to develop local AI tools and solutions.
• Supporting and lobbying with other media actors for better AI policy and regulatory frameworks in the continent - This could take many forms, such as seminars, workshops, and conferences to supporting governments, or working with actors such as media coalitions and the broader tech communities engaged in lobbying.
• Bridging the gender gap - There must be a specific focus on women including through targeted digital training, and media development organisations should support ongoing advocacy work aimed at skilling women with the digital or technical skills necessary for the changing newsroom in Africa.
• Funding research and innovation including the development of local AI tools - This could include initiatives such as the IMS-funded AI Impact Challenge, which sought to support locally developed AI tools that could be used for or to support public interest journalism.