The Natural Language Processing workshop at the Deep Learning Indaba 2025 takes place on, August 22, 2025. It is organised by the Masakhane Community, Data Science Nigeria, Gates Foundation, CLEAR Global and GIZ "FAIR Forward - Artificial Intelligence for All"
Important: This workshop is exclusively for registered attendees of Deep Learning Indaba 2025 in Kigali, Rwanda.
African Natural Language Processing (NLP) has made significant strides in recent years, and new initiatives are already underway to shape the next steps of African NLP while the technology itself is changing rapidly. At the same, key challenges remain: The vast majority of African languages still don’t have models ready for practical applications, resources for compute and data collection while rising remain insufficient and key questions on how to balance opportunities and risks are still open. Against this backdrop and building on last year’s workshop, this workshop will bring together researchers, practitioners, and innovators to assess the current state of African NLP in 2025, explore innovative solutions to common challenges, and foster collaboration and learning to drive the future of African NLP.
Chart a path for advancing African NLP: The workshop will outline the status of African NLP as a basis for charting concrete actions to advance African NLP.
Expected outcomes: Participants will understand where African NLP stands in 2025, and hear about planned and envisioned activities relating to data, evaluation, models and use cases where NLP can create lasting impact.
Understand the opportunities for African innovation for NLP: The workshop will discuss the limitations that African NLP is facing and present innovations to overcome them, such as model distillation or doing NLP in low-resource settings.
Expected outcomes: Participants will discuss and learn from recent advances in areas such as model development in resource-constrained environments or novel ways of data collection for African languages and dialects.
Promote trustworthy and inclusive AI practices: The workshop will present best practices for building inclusive, trustworthy AI applications using NLP and explore how to leverage the opportunities of NLP whilst minimising associated risks.
Expected outcomes: Participants will learn about best practices to develop and use NLP tools for trustworthy AI solutions and explore ethical ways to minimise risks such as extractive data collection or biased models.
Foster cross-disciplinary networks and community-building: The workshop will offer a space for networking, exchange and community-building aimed at unlocking obstacles and advancing African NLP for impact. In particular, it will introduce the new Masakhane AI Hub for African Languages.
Expected outcomes: Participants will engage with speakers, and each other, through interactive formats and explore plans and collaboration opportunities by the new Masakhane AI Hub on African Languages.
08:30–09:00 Registration
10:00 - 10:30 Setting the scene: African NLP in 2025
10:30-11:30 Innovations for African NLP
11.30-12.00 Coffee Break
12:00-13:00 Trustworthy AI for African Languages: Ethical ways for building NLP tools that serve African communities
13:00-13:50 Masakhane Hub: A collaborative effort to advance language AI for one billion Africans
13:50-14:00 Closing Remarks
Confirmed Speakers
Organisers