Journalism can advance trustworthy artificial intelligence in Africa

Technology — and most recently, artificial intelligence (AI) — has revolutionized our information systems. Social media algorithms make assumptions about the type of news or media users want to see, and they influence what is amplified and what is left unseen. To address this large influx of information, AI content moderation enabled by natural language processing has made it easier to detect false or potentially harmful media. Journalists and newsrooms around the world have embraced the technology and incorporated it into how they gather, produce and distribute news. 

However, the development of these technologies is not without blemish. Journalists have played a central role in bringing injustices to the forefront such as the mistreatment of workers who build these models or biases that may be entrenched in them. One cannot ignore the crucial role of journalists and news media in determining whether this technological revolution unfolds positively and in the interest of the people, or if it benefits a small minority at the expense of the larger public. 

Conference panelists discuss journalism excellence in the digital era

This two-fold impact of AI on journalism was explored during theOctober 2023 Conference on the Future of Journalism Education in Southern Africa held in Windhoek, Namibia. The theme of the conference – "Empowering African AI – Enhancing Journalism Excellence in the Digital Era" – emphasized leveraging digital technology to strengthen journalism. It also pointed to the responsibility of African journalists in steering the development and deployment of AI technologies in a direction that empowers Africans. Both obligations require major shifts in media practices and may be a large weight to carry for journalists; however, promoting public good and interest is the main objective of journalism. 

Journalists can advocate for AI regulation in favor of the public interest

There is an urgent need for African governments to develop policies which not only harness the economic benefits of AI but which also respond to the risks of it such as job loss, automated discrimination or the widening digital divide. The policies also need to respond to dependencies which accompany AI such as infrastructural and skills gaps or restrictive foreign policies dictating cross border data flows both within the African continent and with other continents. 

I cannot adequately speak to the technological shifts and adaptations needed to incorporate AI into newsrooms. However, as a researcher critically following policy and governance approaches to data and AI, I have identified a few policy areas where journalists in Africa can strategically apply pressure so that governments set up rules in favor of the public interest. 

Participants at a UNESCO workshop in 1991 that provided the basis for the Windhoek Declaration

The importance of providing the general public with the relevant information they need to meaningfully participate in policy making processes cannot be overstated. In general, press freedom and access to information have improved in most African countries since the 1991 Windhoek Declaration. However, the datafication of information and the introduction of data governance — specifically data protection — may hinder access to information if not implemented correctly. Journalists can be convicted of data breaches if policies do not include special exceptions that protect them. Data protection is vital to curbing state surveillance and the violation of human rights. However, in many data protection laws or bills, uncovering state surveillance or human rights violations may be classified as breaching those laws. Conflicts between access to information and data protection therefore have to be resolved. The public and other civic actors need clarity about certain technologies, their use, and their rules.

Empowering users in the face of big tech 

Providing the public with information is not only essential for their participation in policy making, but it is also essential in bridging the growing power gap between technology companies such as Meta, X (formerly Twitter), or Amazon and their users. Some AI systems are elusive; users' online activities and their personal information are collected to train machine learning models without their consent. There are reports of exploitation of labor, as in the case of Kenyan data annotators, as well as in the case of intellectual property where generative AI is trained on copyrighted data. As more technologies transform industries and create new markets, journalists in Africa will continue to play a critical role in maintaining a power equilibrium between big tech companies and users. Investigating harm caused to people as a result of technologies and telling the stories of those who are being exploited is useful in demanding the reformation of outdated policies which could provide protection or sanction public interest. 

The launch of chatbot ChatGPT in late 2022 has contributed to the hype around AI

Technology is no silver bullet 

Lastly, journalism is paramount in challenging and changing narratives about the capabilities of AI and what policy areas governments should prioritize. The datafication of information and hype around AI has culminated into a situation of techno-solutionism where it is assumed there is a technological fix for every socio-economic issue. There are some problems AI cannot fix — and as a matter of fact there is growing evidence that some of these technologies are exacerbating existing socio-economic injustices. 

Similarly, African governments should not use the AI revolution as a front for their failure to provide and improve basic social services such as education, healthcare, financial and digital inclusion or welfare and safety, to name a few. Critical policy areas which could directly improve the socio-economic conditions of the public or which can address inequality gaps should not be neglected because of the hype around AI. 

The opportunities that AI presents for newsrooms in Africa will no doubt advance journalistic excellence. However, this excellence is predicated on good public policies regarding AI and data, and on a healthy information ecosystem in which journalism is valued. These policies must safeguard human rights and enhance public interest in all spheres of life, including cultural heritage and ensuring environmental sustainability. 

Kristophina Shilongo is a tech policy researcher investigating participatory approaches to AI governance. She is Tech Policy Fellow at Mozilla Foundation, studying participatory governance approaches in the conservation sector in Southern Africa. Her findings will provide  recommendations for AI policies in Southern Africa.

This article was also published on the website of Tech Policy Press and is re-published here with permission.