The Problem We're Solving

In April 2026, South Africa's Department of Communications and Digital Technologies published a national AI policy draft containing fabricated academic citations — generated by AI without verification. The professionals who wrote it were not careless. They simply did not have the framework to know when to question AI output.

This is not an isolated incident. It is the gap that exists across South African workplaces right now:

1.4M
South Africans reached through Microsoft SA’s AI Skills Initiative — trained on digital skills and tool use, not on when to question AI output. (Microsoft, 2026)
37%
Gender penalty found in a 2025 audit of African fintech credit-scoring algorithms — often invisible to the human acting on the recommendation. (Dzreke & Dzreke, 2025; synthetic audit profiles)
5–15%
Completion rate for standard self-paced AI courses. The judgment gap doesn't close through passive learning.

Most existing programmes teach tools. AfriversalAI teaches judgment. While executive AI leadership programmes exist at business schools, no programme yet provides workforce-wide professional credentialing — built around real SA case studies and assessed by human evaluators — for the frontline professionals who encounter AI every day.

Course Philosophy

🌍
African-centred, not imported
Real African contexts — not anonymised US or European scenarios.

POPIA, the Employment Equity Act, the DCDT policy crisis, the 2024 election deepfakes — learners encounter real events from their own professional and civic landscape.

⚖️
Judgment over tools
Transferable critical thinking that outlasts any specific AI tool.

AI tools change rapidly. The judgment skills to evaluate any AI output do not. A learner who completes the course can evaluate any AI system they encounter — including ones that didn't exist when the course was written.

🏅
Demonstrated competency, not completion
The certificate is earned by a real-world assessment, assessed by a human.

Not awarded for watching videos or finishing modules. Earned by submitting a real-world evaluation of an AI system from the learner's own job, assessed against a published rubric. The certificate says "demonstrated," not "completed."

📖
Freire-informed pedagogy
Experience before theory — artifact first, framework second.

Each module begins with a real artifact. Learners engage with it, reflect on their instincts, and then receive the conceptual framework that explains what they were sensing. Learning emerges from the tension between experience and understanding.

🔓
Accessible by design
Zero prior technical knowledge required.

Jargon is introduced only when necessary, always defined in plain language, and grounded in a concrete example. Every concept is explained as if the learner has never formally thought about AI before — because most haven't.

Target Audience & Sectors

Designed for working professionals in five South African sectors — not for AI specialists, data scientists, or software engineers, but for the colleagues they work alongside.

🏢
Corporate & Business
🏛
Government & Public Sector
💰
Finance & Banking
🏥
Medical & Healthcare
🎓
Education
Layers 1 & 2 — Primary Learner
  • Working professional, 5+ years in their field
  • Already encountering AI tools at work
  • Not a developer — doesn't write code or train models
  • Accountable for decisions influenced by AI recommendations
  • Has not received formal AI literacy training
Layer 3 — Professional Learner
  • Policy analyst, compliance officer, legal counsel, or regulator
  • Directly shapes, oversees, or enforces AI-related decisions
  • Needs specialist credentials to operate at a professional standard
  • Layer 1 Core Certification required as prerequisite
Organisational buyers (L&D and HR): Corporate and government L&D teams purchasing workforce training. The course is designed to be MICT SETA-accreditable, enabling employers to recover 40–60% of training costs through the Skills Development Levy discretionary grant system.