The study “AI Adoption in South African Newsrooms: Exploring Journalists’ Perceptions” by Soligah Solomons and Musawenkosi W. Ndlovu both from University of Cape Town, looked at a topic we have also previously written about from a closely related perspective: AI adoption in South African newsrooms and how journalists perceive it.
The study was qualitative in nature and relied on 17 in-depth interviews with journalists from South African news organizations. They were asked, among other things, whether they already used AI, their impressions of it, if newsrooms’ adoption of AI could increase the number of stories being produced, if it could give journalists more time to investigate stories, whether there are inherent biases in AI systems that they think may contribute to perpetuating societal inequalities, and whether the deployment of AI could facilitate inclusion of more marginalised and local voices in mainstream news coverage.
The journalists viewed AI as inevitable in the newsrooms. This view is congruent with the scholarly discussion on the topic, such as Hassan and Albayari (2022). One journalist described it as a new buzzword that comes up in discussions about other matters.
However, AI is still in its developmental stage in South African newsrooms, and the discussions center on ethics and guidelines. Newsrooms are still working on policies to regulate the use of AI and its impact. However, there was a notable lack of internal deliberation among journalists at publicly owned media.
As the adoption of AI is inconsistent, the authors felt it necessary to distinguish between “formal newsroom-sanctioned AI use by journalists and the unofficial adoption of AI by individual journalists”, as they put it. Some noted that they had experimented with it, but that it is not part of everyday processes. Few South African newsrooms had official policies for it.
There is a gap between the Global North and South African unique circumstances. Despite the advantages of AI integration in the North (as noted by Marconi 2020; Moran and Shaikh 2022), many journalists in South Africa noted that there are no plans to introduce AI for news reporters, and many are not empowered to use AI. The authors argue that financial constraints and technical resource shortages account for the disparity.
Journalists acknowledged the usefulness of AI in the quantity of articles produced, but many still felt that human involvement was necessary. For quality, AI was found not to be correlated with investigative journalism, with authors agreeing with Stray’s (2019) conclusion that while AI can automate some tasks, journalists remain essential for the most part.
When it comes to inherent biases, journalists were concerned that AI might perpetuate existing biases in the newsrooms. One stated that it might amplify existing social biases if care was not taken in how it is implemented. The use of AI in South Africa must be carefully assessed (Gondwe 2023; Munoriyarwa et
al. 2023).
The journalists were particularly pessimistic about the ability of AI to foster local voices, as by mining data that already exists, it excludes marginalized communities and voices. There were also concerns about dependence on the AI companies’ platforms and the exclusion that comes with it. The authors suggested that South African newsrooms must develop AI technologies aligned with African
journalistic principles, if local voices are to be heard and expressed.
In conclusion, the paper added to the understanding of AI adoption in the South African context, and further highlighted the differences of the African context to the situation in developed countries in the Global North.
The article “AI Adoption in South African Newsrooms: Exploring Journalists’ Perceptions” by Soligah Solomons and Musawenkosi W. Ndlovu is in Communicatio. (Open access).
Picture: Clifton beach (Cape Town) by kylefromthenorth.
License Unpslash.
