Looking Back: 30 Years of Promise, 18 Months of Impact 🌍

Two weeks ago, we gathered at the University of Pretoria for our 2026 AI4D African Languages Lab Workshop.

It was one of those rare moments where the academic, the legal, the technical, and the community-based worlds collided in the best possible way. We spent the day grappling with a heavy question: After 30 years of constitutional promises of language equality in South Africa, why is the gap between policy and practice still so wide?

Thanks for reading DS@UP 🚀! Subscribe for free to receive new posts and support my work.

It was a day of vulnerability, “uncomfortable conversations,” and, most importantly, a collective pivot toward action.

A Recap of the Day

From the morning framing by Prof. Vukosi Marivate to the intense panel discussions on “The Content Pipeline,” we moved beyond theoretical research. We interrogated the “last mile” of development—how we move from an experimental language model to a deployed system that doesn’t just work in a lab, but actually serves the people in Limpopo, the farmers in the field, and the patients in our hospitals.

Key highlights include:

• The Content Pipeline: Insights from SADiLaR and Nthavela Community Media on why translation is about preserving meaning, not just words.
• Resilient AI: A masterclass by Prof. Hein Venter on why, if the first era of AI was defined by scale, the next must be defined by resilience and forensic accountability.
• Ethical Guardrails: A deep dive into the legal complexities of product liability, copyright, and why we need an “Afrocentric” regulatory framework for AI.
• Our Anniversary Celebration: We wrapped up our evening reception at Pure Cafe by reflecting on our first 18 months—a whirlwind of capacity building, cross-continental collaboration, and the birth of a new research institute (AfriDSAI).

Why This Matters Now

We are at an inflection point. As AI becomes the bedrock of societal infrastructure, we face the threat of “Model Collapse”—where synthetic data degrades our models’ ability to understand the nuance of our diverse languages.

We heard loud and clear from our speakers: “People don’t stay without care.” Building African language AI is not just about compute power; it is about stewardship. We are committed to building an ecosystem where African languages aren’t just “supported” but are fundamental to the digital world.

Catch Up on the Conversations

Whether you missed the sessions or want to revisit the insights from our panelists and technical leads, we have curated everything into one place.

👉 Read the Full Workshop Recap & Watch the Sessions

What’s inside the recap:

• Full Session Recordings: A complete YouTube playlist.
• Community Milestones: Insights from our Hundzula retreats and SWIP Wikipedia collaborations.
• Research Tools: Access to our playbook on the Digitisation of Oral Data for NLP of Low-Resource Languages.

The Road Ahead

The energy in the room was a call to action. We’re moving forward with a new national task team and a roadmap that prioritizes interdisciplinary, “human-in-the-loop” AI.

If you are a researcher, a student, or an industry partner, we want you in the ring with us. Connect with the DSFSI Lab on LinkedIn and follow AfriDSAI for the latest updates.

A huge thank you to the International Development Research Centre (IDRC) and the UK’s Foreign, Commonwealth and Development Office (FCDO) for making this community possible.

Let’s keep building.

Thanks for reading DS@UP 🚀! Subscribe for free to receive new posts and support my work.

Stay connected with our work:

• DSFSI: https://linktr.ee/dsfsi
• AfriDSAI: https://linktr.ee/afridsai