The workshop is structured to explore both the human-centric and technical aspects of AI, particularly in under-resourced communities within the Global South. It will delve into critical considerations such as data generation, annotation, legal frameworks, and sharing protocols, emphasising culturally contextual datasets and the governance frameworks that support them. Furthermore, the workshop will address sharing AI models and datasets through various licensing frameworks, emphasizing ethical data collection, community engagement, and the management of AI innovations in a manner that respects local cultures and languages.

Abstract

The transformative potential of AI and ML is immense, reshaping sectors from healthcare to finance. However, the pace of AI development and its benefits are not uniformly distributed, with under-resourced communities, particularly those historically disadvantaged, experiencing a lack of access to the rewards of AI. This disparity highlights the need for responsible, ethical, and community-centric approaches to AI development that prioritize local needs, cultures, and languages. This workshop aims to address these disparities by focusing on the human aspects of AI. It seeks to move beyond the technicalities of algorithms and computation to delve into the foundational elements of AI, the data itself. Doing so emphasizes the importance of data practices that are efficient, ethical, culturally aware, and inclusive. Participants will engage in discussions and practical sessions that explore data generation, annotation, legal frameworks, and sharing protocols. Special attention will be given to creating culturally contextual datasets that reflect African communities’ linguistic and societal nuisances. The workshop will also cover critical topics such as data governance, synthetic data implications, and cross-cultural data safety challenges and opportunities. The goal is to equip researchers, ML practitioners, and policymakers with the knowledge and skills necessary to implement AI projects that are ethical, inclusive, and impactful, thereby fostering an AI ecosystem that truly benefits all, especially those in marginalized communities. Through this endeavour, the workshop will contribute to shaping a future where AI development in Africa is driven by African voices and directly addresses the continent’s unique challenges and aspirations.

Target Audience

This workshop is designed for AI researchers and enthusiasts, ML practitioners, data scientists, and legal experts engaged in or interested in responsible AI development within African contexts. It will also be valuable for stakeholders from global data training companies, open-source data benchmark consortia, and technology firms operating in Africa.

Workshop Objectives & Goals

1. To equip participants with skills in efficient AI data collection and ethical data utilization.
2. To discuss and provide recommendations on proprietary, open, and open-source licensing for sharing NLP datasets and AI models.
3. To foster a deeper understanding of AI policy, data rights, and the ethical use of AI in Africa.
4. To engage participants in creating a multilingual, augmented dataset that reflects diverse African languages and contexts.
5. To facilitate discussions on building sustainable economies with data and leveraging synthetic data creation.

Detailed Full-Day Workshop Agenda

TBD

Organizing Committee

• Chris Emezue, Lanfrica Labs
• Dr. Chijioke Okorie, Data Science Law Lab, University of Pretoria
• Dr. Sarah Luger, ML Commons and Consumer Reports
• Dr. Melissa Omino, Centre for Intellectual Property and IT Law, Strathmore University
• Jade Newton, NLP Data and Business Consultant
• Florence Ogonjo, Centre for Intellectual Property and IT Law, Strathmore University
• Catriona Anyango, Centre for Intellectual Property and IT Law, Strathmore University