By: Ahmed Fathi
New York: The African Union (AU) is taking bold steps to address the multifaceted challenges posed by the rapid rise of artificial intelligence (AI), a technology that promises to reshape economies and societies across the globe. Through its development arm, AUDA-NEPAD, the AU is diligently working to forge strategic partnerships with major U.S. technology firms while simultaneously crafting a comprehensive continental policy aimed at addressing critical issues such as infrastructure development, energy capacity, and data sovereignty.
At the core of the AU’s strategy lies the daunting task of developing an AI ecosystem that is truly African in character. AUDA-NEPAD, under the leadership of Nardos Bekele-Thomas from Ethiopia, has become the driving force behind this vision. Operating from its headquarters in Johannesburg, the agency is actively exploring ways to build the necessary infrastructure to support AI innovation across the continent. This includes ambitious plans to enhance supercomputing capabilities and establish data centers in regions that currently lack such facilities.
However, the challenges facing the AU are immense. Africa’s current computing power is alarmingly insufficient, with only a few countries like South Africa, Kenya, Senegal, Ivory Coast, Morocco and Egypt. possessing supercomputers capable of handling the vast computational demands of modern AI. Central Africa, in particular, is conspicuously absent from this technological map, lacking even a single supercomputing facility. In response, AUDA-NEPAD is considering a regional approach to infrastructure development, proposing the creation of a data center in Gabon that could serve multiple countries. Yet, this strategy is fraught with difficulties, not least of which is the significant financial investment required to make such a project viable.
Another critical element of the AU’s approach is its engagement with powerful Western technology companies. For example, AUDA-NEPAD has been in discussions with Google and has already signed a memorandum of understanding with Meta, the parent company of Facebook. These partnerships are intended to provide African engineers with essential training in AI, offer access to vital metadata, and establish centers of excellence on the continent. While these initiatives are undeniably valuable, they also raise serious concerns about Africa’s data sovereignty and the potential for a new form of digital colonialism, where African nations could find themselves heavily dependent on foreign entities for their technological advancement.
The energy infrastructure of the continent presents yet another formidable obstacle. Africa’s total installed electrical capacity, which stood at 245 gigawatts in 2022, is woefully inadequate to meet the enormous energy demands required by AI technologies such as GPT-4. Without a dramatic increase in energy production, the AU’s aspirations for AI leadership may be stymied by the simple reality that the continent lacks the power necessary to fuel these advanced systems.
The AU’s reliance on external partnerships and funding also comes with its own set of risks. Public-private partnerships (PPPs) and external financing are being considered as potential solutions to bridge the gaps in infrastructure and energy. The European Union, for instance, has shown interest in contributing to these efforts through its Global Gateway strategy, which may include financing for data centers and supercomputers. However, the use of PPPs, such as the one Cameroon has with Huawei for its government datacenter, could lead to a loss of national control over critical digital infrastructure, thereby compromising state sovereignty.
The AU’s focus on Western technology companies, while understandable given the current landscape, may also inadvertently reinforce Africa’s dependency on external powers. Notably, the upcoming Forum on China-Africa Cooperation (FOCAC) in Beijing does not feature any AI-related discussions on its agenda, reflecting the limitations of non-Western actors in this technological sphere. This omission highlights a broader issue: as Africa seeks to develop its AI capabilities, it may find itself increasingly reliant on Western partners, at the risk of being left out of alternative global tech alliances.
As the AU advances its AI strategy, it faces a delicate balancing act. The organization must navigate the complexities of building indigenous AI capabilities while managing external dependencies and safeguarding the sovereignty of its member states. The struggle over Africa’s digital future and its role in the global AI landscape is just beginning, and the decisions made by the AU in the coming years will have profound implications for the continent’s place in the emerging digital world. The stakes could not be higher, and the challenges are as formidable as the opportunities that lie ahead.
Comments