The increasing incorporation of AI into everyday activity is a source of excitement for many as it is set to improve productivity and innovation across sectors — healthcare, finance, sports, and even politics; the list is endless. In a country like Nigeria where young people have leveraged social media platforms to drive political actions like #EndSARS and #BringBackOurGirls, and where politicians have used social media platforms and citizens-facing techs for electioneering and political engagement, it is no surprise that AI is increasingly featuring in its political landscape. 

But AI also bears negative consequences, as evident in the 2019 general elections in the country in which generative AI tools were used to manipulate public opinion, exacerbating political divisions and undermining the most important political process in any democracy. This reality is a source of concern because if AI, which is still in its experimental phase in Nigeria, already portends pockets of significant concerns to our democracy, then its inevitable growth will only make things worse. 

This becomes even more concerning because this misuse extends beyond individuals to state actors and institutions as AI increases the possibility of mass state surveillance, targeted disinformation, and political propaganda, raising concerns about human rights violations and the erosion of civic spaces. These realities were the central focus of a recent meeting to promote AI accountability in Nigeria, among media and civil society organisations (CSOs) stakeholders. Hosted by the  Joint Civic Defense Fund (JCDF) and Luminate in Abuja last month, the convening on ‘Artificial Intelligence, New and Emerging Technologies, and Closing Civic Spaces’ featured prominent organisations, including Dataphyte’s Goloka.

At the core of the issues examined was an evaluation of the impact of AI on the civic space in the country and how civil society organisations (CSOs) and the media currently utilise AI and emerging technologies to counteract the existing restriction of civic spaces and the potentials to leverage it for the advancement of democracy and development.

Similarly, stakeholders evaluated the acceptance and progress of AI efforts towards enabling inclusiveness, ensuring women, minority groups, and persons with disabilities actively engage its tools to improve civic participation and avoid further marginalisation.

Importantly, the convening noted that the media has a critical role to play in enabling people to positively engage AI and leverage its tools for development. This calls for media organisations to leverage their role as custodians of information to enhance the quality of discussions around the technology and to educate the public and bridge the current access to information gap. But this effort cannot happen in isolation; rather than waiting to form reactive response alliances once a negative AI use case goes viral, there is a need for proactive collaboration between the media and civil society organisations to educate the public on AI, empower the civic space to use it, and contribute to a robust AI ecosystem in general. 

Another key issue is the scope AI regulation should take. Stakeholders highlighted the need for the National Strategy in progress to emphasise the responsible use of AI and other emerging technologies in the country, a critical step to creating a national direction on its application. 

Furthermore, ensuring that AI tools are designed or adapted and deployed with sensitivity to the Nigeria social context is very important to ensure debiased and culturally representative outputs. The issue of data representativeness and AI resourcing in African languages is a key issue that many African stakeholders are currently grappling with.

Generally, most AI solutions in Nigeria are either at the design or experimental stage, and many of these struggle with sourcing appropriate data. Even when they have access to a wide pool of datasets and resources to refine their AI models, the question lies in the acquisition of this data which often overlooks the existing diversity and nuance of each environment, especially in developing countries. 

To this end, the convening encouraged software developers to create AI platforms that cater to varied groups, languages, and demographics. Goloka by Dataphyte emerges as a major platform addressing the AI data collection challenges in Nigeria.

Designed as a one-stop shop for data collection and research design, Goloka gathers, labels and sorts diverse and dispersed datasets covering images, indigenous language voices, and textual data on hard-to-reach groups to help governments, civil society organisations, and SMEs make more informed and inclusive decisions. The platform uses a network of verified field agents to source and curate diverse datasets efficiently, transparently, and ethically. This enables it to provide the diverse datasets collected required to refine AI models with a higher degree of accuracy.