Just AI in Africa

The rapid growth of AI technology around the world holds the potential to improve research and innovation, but also the risk of aggravating inequality in access to information and economic opportunities. A key problem is that most of the small number of critical “foundation” and “frontier” models upon which AI applications are built are created and controlled by an even smaller number of multinational commercial technology firms. These models are rarely transparent, often culturally and linguistically biased, and have immense market power. One response to the promise and challenges, especially in the Global South and the so-called “middle powers,” is the promotion of “Public AI” — referring to government led and/or publiAt the Centre on Knowledge Governance, in consultation with other key players, we are developing a somewhat broader conception of “Just AI” that blends the desires for public accountability in Public AI with human rights concerns, including the moral and material interests of creators, the stewardship of traditional knowledge by communities and the developmental priorities of the Global South. The purpose of this series of roundtable discussions followed by a drafting retreat in the Cradle of Humankind is to produce a “zero draft” statement of experts on principles of Just AI for Africa.