The study “Artificial Intelligence Practices in Everyday News Production: The Case of South Africa’s Mainstream Newsrooms” by Allen Munoriyarwa, Sarah Chiumbu and Gilbert Motsaathebe, all from University of Johannesburg, looks at to what extent artificial intelligence (AI) has been used in South African newsrooms and how journalists and editors perceive it.

There is no common understanding on how to define artificial intelligence used in journalism. Jamil (2020) sees automated news as part of AI, but Broussard (2019) rejects this, for example. The authors of this article seem to side with Jamil, using a broader definition of AI. 

Nevertheless, the use of AI in journalism can be traced to early 2000s, when leading news organizations Reuters and New York Times began using algorithms and automated news writing in their journalism. In Africa, the use of AI in newsrooms is relatively new. 

The South African media environment is characterised by plurality, diversity, vibrancy and liberality, in addition to multilingualism. Recently, in the past four years, changes in the economic environment to a more unstable one have led to mergers, job losses, and closings of some media organizations. 

Bigger media actors have recently started to utilize more technology-driven approaches like data journalism, but no previous literature exists on the adoption of AI in South Africa. AI in journalism elsewhere has been researched, with the debate being characterized by a continuum of optimism, pessimism, and cautious optimism.

The authors conducted semi-structured interviews of journalists among six media organizations (3 television stations, 2 newspapers and online news outlet) in South Africa that were at various stages of technology adoption. Or, as in the framework of the study, technology appropriation.

The questions in the interviews sought to understand the extent to which AI practices had been adopted and how they had changed the reporting practices. Also, the effect to the journalists’ contribution to democratic processes and the quality of journalism affected were asked about. 

The overall adoption of AI in South Africa has been slow, with only the biggest media organizations including it to an extent. The interviewees were generally pessimistic about AI, with some viewing it as “Western practices” that have no place in African journalism. 

The reasons for the pessimism were four: first, there was a belief that there were not enough skills among the journalists in South Africa to understand the new technological processes. Decond, the newsrooms lacked financial resources to implement AI-driven newsmaking.  

Third, many were concerned about the ethical dilemmas surrounding AI journalism. Fourth, many respondents felt that AI-driven practices did not contribute to media’s Fourth Estate role in democracy.

In contrast to Europe and US, where AI-driven journalism has become a standard practice, newsrooms elsewhere do not yet use it to the same extent. The results therefore mirror the ones from other non-Western countries. 

However, unlike other factors, in South Africa, the availability of AI technology and the journalists ability to appropriate it are the important factors. AI appropriation is also directly linked to accountability and contribution to democracy, where the concerns lie. 

The authors argue that the pessimism concerning the role in democracy and job losses are well-founded in the context of developing economies, and should not be sidestepped. They argue that journalists should retain total control over news production.

It is also possible that AI might not, like it does in the West to an extent, restore the credibility of journalism in developing countries. Instead, human-driven news production that supports democratic processes should take priority.

The article “Artificial Intelligence Practices in Everyday News Production: The Case of South Africa’s Mainstream Newsrooms” by Allen Munoriyarwa, Sarah Chiumbu and Gilbert Motsaathebe is in Journalism Practice. (free abstract).

Picture: Cities have their own mystique. Dynamic, Chaotic, Alive. By Jacques Nel

@jcqspht.

License Unsplash.