Layer 1 · Core Curriculum

Core Modules 0 – 6

The foundation every EmriVersal Ai learner completes, regardless of sector or specialization. Module 0 and six core modules, all built around real, documented AI cases.

Four-phase structure — every module: Encounter (~20 min) → Reflect (~20 min) → Concept (~30 min) → Apply (~30 min). Approximately 100 minutes per module self-paced. Facilitated sessions run 90–120 min with peer discussion.
0
AI Fundamentals: What Is It, Really?
Free · All sectors · Required pre-work before cohort
Completed before the cohort begins — free, no registration required. Builds the shared vocabulary all sector tracks depend on. ~45 minutes self-paced.
Learners will be able to:
  • Define AI in plain language and distinguish it from other software
  • Identify the three types of AI encountered in professional settings (Generative, Predictive, Computer Vision)
  • Name at least five things AI cannot do, regardless of how confident it sounds
  • Spot one AI system in their own workplace and classify it by type
Concept3 types of AI, 5 things it cannot do
EncounterSort 10 everyday tools: AI or not AI?
ReflectWhere is AI already in your job?
ApplyClassify one workplace tool & name a failure mode
1
The Confidence Problem
Free preview available
📄 AI hallucination in a national policy draft · South Africa, April 2026

What makes AI different from other software? What can generative AI not do — regardless of how confident it sounds? Why does it hallucinate sources and fabricate statistics?

Learners will be able to:
  • Explain how generative AI produces output without "knowing" anything
  • Identify the key failure modes of large language models
  • Spot hallucination signals in a real document (a national policy draft, South Africa)
EncounterRead the policy excerpt — no context given
ReflectWhat made you trust or distrust it?
ConceptAI hallucination explained via the real scandal
ApplyApply the Funda Five™ to that policy scenario
2
How AI Learns — and Fails
💰 African Fintech Credit Scoring Bias · 2025

AI is trained on data. That data reflects the world's inequalities. Understanding how AI learns is the first step to spotting when it's failing people — including you.

Learners will be able to:
  • Describe how AI learns from historical data in non-technical terms
  • Explain why training data reproduces systemic bias
  • Identify proxy discrimination in a real AI credit-scoring scenario
EncounterA credit decision output — something is off
ReflectWho does this outcome affect?
ConceptTraining data, proxy variables, feedback loops
ApplyAudit your organisation's data inputs
3
Bias and Discrimination
🔍 When a neutral-looking rule lands harder on one group

Discrimination can happen without discriminatory intent. In the UAE, AI trained on historical data can carry forward patterns tied to nationality, residency status and length of local record — often invisibly.

Learners will be able to:
  • Define algorithmic bias in plain language for a non-technical audience
  • Analyse an AI recommendation for disparate impact on a protected group
  • Apply an equity lens to AI deployment decisions in their sector
EncounterPolicing scenario — who gets flagged?
ReflectWhat assumptions are built into this system?
ConceptDisparate impact, proxy discrimination, UAE law
ApplyEquity audit of a tool in your sector
4
Verification and Judgment
📱 Article 44 — manipulated media and intent to defame

Verification is a professional skill, not a technical one. How do you decide when AI output is good enough to act on — and when it needs a second look?

Learners will be able to:
  • Apply the Funda Five™ Trust step to any AI output
  • Identify deepfake and synthetic content signals in audio and video
  • Build a personal verification checklist relevant to their role
EncounterA campaign clip — real or synthetic?
ReflectHow would you verify before sharing?
ConceptDeepfake techniques and verification methods
ApplyBuild your personal AI verification checklist
5
Accountability and Governance
⚖ AI Recruitment + the UAE Labour Law (Federal Decree-Law No. 33 of 2021) Liability · 2025

"The algorithm decided" is not a legal or ethical defence under UAE law. When AI affects someone's employment, credit, or healthcare, a person in an organisation is accountable.

Learners will be able to:
  • Map the accountability chain for an AI decision in their own organisation
  • Identify the UAE Labour Law (Federal Decree-Law No. 33 of 2021) exposure from AI recruitment tools
  • Raise a substantive AI concern through the right organisational channels
EncounterCandidate rejected by AI screener
ReflectWho in your org made this decision?
ConceptArticle 4 of the UAE Labour Law, the Personal Data Protection Law, where accountability sits in law
ApplyMap your org's AI accountability chain
6
When AI Works — and When It Doesn't
✅ Where AI helps, and where a person must decide

The question is never "AI yes or no?" It's "AI for what purpose, for whom, in what context, with what oversight?" This module synthesises the entire course through two contrasting cases.

Learners will be able to:
  • Evaluate an AI deployment across all five Funda Five™ dimensions
  • Write a professional AI judgment statement suitable for a team briefing
  • Articulate when AI deployment is and isn't appropriate in their sector
EncounterTwo AI systems — same goal, different design
ReflectWhat makes one trustworthy and one not?
ConceptFull Funda Five™ applied to both cases
ApplyWrite your personal AI judgment statement
After Module 6, every learner completes one sector-specific module — see Sector Tracks →
← Funda Five™ Next: Sector Tracks →