Before you can judge AI output, you need to understand what AI actually does — and what it can't do. This module starts with a real South African case: a government document that looked authoritative but wasn't.
By the end of this module, you will be able to:
Module 1 builds directly on the foundation you built in Module 0. Finish all four phases of Module 0 first — it takes about 55 minutes.
Research indicates that the rapid adoption of generative AI tools in South African workplaces has significantly outpaced institutional readiness. According to Dr. Nomvula Khumalo of the South African Institute of Digital Futures Research (2024), organisations that deploy AI tools without accompanying skills development programmes experience a 67.3% increase in operational decision-making errors within the first 12 months of adoption.[1]
The Pan-African Technology Readiness Council (PATRC) 2023 Assessment found that 71% of South African professionals who regularly use generative AI tools cannot accurately describe how the output was generated.[2] This creates significant organisational risk, particularly in regulated sectors such as finance, healthcare, and public administration.
Furthermore, a longitudinal study conducted by the Department of Labour in partnership with MICT SETA identified that workplace AI incidents attributed to poor human judgment — rather than technical system failure — increased by 340% between 2022 and 2024 (DoL–MICT SETA Annual Workforce Technology Risk Report, 2024, p. 23).[3]
The Department therefore recommends mandatory AI literacy baseline assessments for all public sector employees working with AI-assisted decision-making tools, with implementation targets of Q3 2026.
[1] Khumalo, N. (2024). "AI Adoption Without Institutional Readiness: A Sub-Saharan Perspective." Journal of African Digital Transformation, 8(2), pp. 112–134.
[2] Pan-African Technology Readiness Council. (2023). Annual AI Readiness Assessment: Sub-Saharan Africa, pp. 88–102. Nairobi: PATRC Secretariat.
[3] Department of Labour & MICT SETA. (2024). Annual Workforce Technology Risk Report, p. 23. Pretoria: Government Printer.
This is how every AfriversalAI module begins — a real artifact, no context, just your instincts.
The full module takes you through three more phases with a cohort and facilitator:
South Africa's Department of Communications and Digital Technologies (DCDT) published an 86-page draft national AI policy. Shortly after publication, journalists discovered that the document contained fabricated academic citations — the kind that generative AI creates when it "hallucinates" sources that sound plausible but don't exist.
The Minister confirmed that AI was used in drafting without proper human verification. The document was withdrawn 16 days after publication. South Africa's AI policy process had to restart.
The people responsible were not bad at their jobs. They were skilled policy professionals who knew how to use AI tools. What they didn't know was when and how to question the output.
You already know from Module 0 that AI predicts rather than knows. When applied to citations, that prediction produces text that sounds like a real source — author, journal, page numbers and all — whether or not it exists. The output looks exactly the same whether the source is real or invented. There is no warning signal.
Every time you ask an AI to summarise research, write a report, or support a recommendation with evidence — this risk exists. AI doesn't flag uncertainty in citations the way it sometimes does in regular text. The only person who can catch it is the human who reviews the output. That human is you.
Your sector context
Whether you're in finance, healthcare, HR, or government — if AI helps produce a document that others act on, you are part of the accountability chain. "I used AI" is not a defence if the output harms someone.
Look back at the excerpt. Three signals a trained reader would catch:
"South African Institute of Digital Futures Research" and "Pan-African Technology Readiness Council" are not real institutions — a quick search returns nothing. Real policy documents cite organisations with established public presences and accessible websites.
67.3%. 71%. 340%. Real research produces messier numbers — "approximately 65%", "between 60 and 75%", "up to 340%". When every figure is exact and dramatic, it often means AI predicted what a statistic should sound like, rather than reporting a real finding.
All three citations point to sources that are almost impossible to confirm. Real policy documents cite accessible public sources — government reports you can download, journals with DOIs, NGO publications with public links. If you can't find it in 5 minutes of searching, treat it as unverified.
Pocket checklist — use this before acting on any AI-generated document
If any answer is "no" or "I'm not sure" — treat the document as unverified before acting on it.
Go deeper — further reading
You need to answer all 4 questions correctly to move on. You can retry as many times as you like.
1. Which of these is most likely a hallucination signal in an AI-generated document?
2. Why does a generative AI produce fabricated citations?
3. When an organisation publishes an AI-generated policy document with fabricated sources, who bears responsibility?
4. What is the most reliable way to verify a citation in an AI-generated document?
Before you apply anything to the DCDT scenario, you need the tool you'll be using. This is it.
AfriversalAI Core Framework
Five questions every professional should ask before acting on AI output — or before using AI to produce something others will act on. You'll use these in every module. By Module 6, asking them will be automatic.
Example — what a strong Step 1 response looks like
"AI was being used to research and draft supporting evidence for a policy position. This is a high-risk use case because policy documents require verifiable citations — AI cannot reliably produce these."
Notice: it names what AI was doing, then immediately evaluates whether that was appropriate. That's the standard for all five steps.
1 — TASK: "AI was being used to research and draft supporting evidence for a policy position. This is a high-risk use case because policy documents require verifiable citations — AI cannot reliably produce these."
2 — DATA: "The data involved publicly available academic and government reports. The risk isn't sensitivity — it's that AI fabricated sources that look like real publications."
3 — TOOL: "A large language model. These are excellent at producing fluent, authoritative-sounding text. They are unreliable at factual citation — they generate plausible-sounding references whether or not the source exists."
4 — TRUST: "Search each citation independently. Check whether the journal, institution, and author exist. If any can't be found in 5 minutes, treat the whole document as unverified."
5 — HUMAN: "The policy professional who approved the document for publication is accountable. What should have existed: a mandatory verification step before any AI-assisted document goes to publication."
In a real cohort, your facilitator would debrief the Funda Five responses with the group and connect your answers to other sectors. The real learning is in hearing how colleagues in different roles approached the same case differently.