rudi Start Here
Menu

Learning and development

AI Literacy Framework Matrix

Understand AI capability across three dimensions: Conceptual, Operational and Governance. The matrix describes five proficiency levels through both nontechnical and technical applications.

Download PDF3 pages · PDF · November 2026

Conceptual

Understanding AI at a high level. How AI works, why outputs vary, and the limitations and societal implications of its use.

Three dimensions · Five proficiency levels · Two application tracks

ProficiencyNontechnical applicationTechnical application
LEVEL 0No ExposureHas never heard of generative AI.Has no awareness of AI/ML technologies.
LEVEL 1Pre-BeginnerHas heard the term “AI,” maybe ChatGPT, but cannot explain it.Knows AI exists but confuses it with automation; no exposure to models.
LEVEL 2BeginnerCan describe AI as “a tool that generates text/images.”Knows AI involves training data and models; aware of major model providers.
LEVEL 3IntermediateUnderstands basics: training data, bias, why outputs vary.Understands concepts like parameters, embeddings, fine-tuning; can explain architectures in plain language.
LEVEL 4AdvancedCan critically discuss AI’s role in society, ethics, and limitations.Can explain trade-offs in model design, architectures, and evaluation metrics.
Read each dimension separately. Someone may be experienced in one lane and new to another. Use the descriptors to identify the next learning objective; the full matrix contains 30 competency descriptors.

Operational

Using AI tools in practice. The work a person can complete with AI, from a first prompt to a repeatable workflow.

Three dimensions · Five proficiency levels · Two application tracks

ProficiencyNontechnical applicationTechnical application
LEVEL 0No ExposureHas never opened an AI tool.Has never run an AI model, API, or notebook.
LEVEL 1Pre-BeginnerCan open ChatGPT/DALL·E but doesn’t know what to type.Has installed Python/JS but hasn’t touched AI libraries.
LEVEL 2BeginnerCan try basic prompts (“Write an email…”), recognizes when output is off.Can run a pre-built model or call a basic API.
LEVEL 3IntermediateUses AI for work tasks: lesson plans, emails, reports; knows how to cross-check.Builds small automations (APIs, n8n/Zapier), fine-tunes models with guidance.
LEVEL 4AdvancedDesigns AI-enhanced workflows, mentors peers in safe usage.Builds custom pipelines, deploys apps, integrates multiple models, and optimizes performance.
Read each dimension separately. Someone may be experienced in one lane and new to another. Use the descriptors to identify the next learning objective; the full matrix contains 30 competency descriptors.

Governance

Oversight, ethics and responsible use. How a person recognizes risks, applies policy and supports safe AI adoption.

Three dimensions · Five proficiency levels · Two application tracks

ProficiencyNontechnical applicationTechnical application
LEVEL 0No ExposureHas no awareness of AI risks.Has no awareness of AI ethics or regulation.
LEVEL 1Pre-BeginnerKnows AI may be biased, but no idea how.Has heard of bias, fairness, or GDPR but never engaged.
LEVEL 2BeginnerCan name risks: privacy, misinformation, bias.Knows about data handling, PII, responsible disclosure.
LEVEL 3IntermediateKnows organizational policies, applies them (e.g., not inputting student data).Implements policies: data filters, logging, audits.
LEVEL 4AdvancedLeads conversations on AI policy, ethics, and adoption.Designs governance frameworks, bias testing, compliance systems, transparency reports.
Read each dimension separately. Someone may be experienced in one lane and new to another. Use the descriptors to identify the next learning objective; the full matrix contains 30 competency descriptors.