🤖
About this framework
This page maps what teachers should be able to do with AI, across 5 areas of work and 3 steps (Foundation → Practitioner → Leader). Click any box in the matrix below to read the rule, what to try this week, what to avoid, and how to grow.
For · all teachers
Read-only reference
Eduversal AICF v1.0 · Part 1
Approved · Academic Directorate
Verbatim framework text lives in docs/research/eduversal/ai-competency-framework/ and is surfaced read-only on /references?doc=aicf-* — this page never edits it.
How to use this page
Every teacher should reach Foundation in the first year. Most teachers reach Practitioner after that. Leader is for teachers who help others — not everyone has to get there.
Click a box to read the exact words from the framework. The same window also shows what to try this week, what to never do, how to move to the next step, and which other pages to read.
Competency Matrix
3 Levels × 5 Domains
Foundation
Practitioner
Leader
Domain A
Knowledge & Understanding
Knowledge & Understanding
Understands AI basics, key limitations (hallucination, bias), and ethical considerations. Knows the school's AI policy.
Foundation · A
teacher.foundation.domainA
Deeper understanding of how LLMs work. Stays current with developments. Can explain AI to students and parents.
Practitioner · A
teacher.practitioner.domainA
Comprehensive understanding. Engages with technical concepts and research literature. Anticipates future developments.
Leader · A
teacher.leader.domainA
Domain B
Practical Application
Practical Application
Uses approved tools for lesson planning, materials creation, admin tasks. Reviews AI outputs before use. Follows data privacy.
Foundation · B
teacher.foundation.domainB
Advanced prompt engineering. Uses AI across differentiation, rubrics, feedback templates. Troubleshoots when outputs aren't right.
Practitioner · B
teacher.practitioner.domainB
Expert-level skills. Designs complex AI-enhanced workflows. Evaluates tools for school adoption. Develops resources for colleagues.
Leader · B
teacher.leader.domainB
Domain C
Pedagogical Integration
Pedagogical Integration
Uses AI materials to support learning objectives. Remains the instructional decision-maker. Can explain AI use to students.
Foundation · C
teacher.foundation.domainC
Integrates AI strategically for differentiation. Designs learning that leverages AI while maintaining rigour. Teaches students responsible use.
Practitioner · C
teacher.practitioner.domainC
Designs innovative AI-enhanced learning. Develops curriculum materials for the school/network. Mentors colleagues. Leads student AI literacy work.
Leader · C
teacher.leader.domainC
Domain D
Critical Evaluation
Critical Evaluation
Fact-checks AI content. Spots obvious errors, false citations, cultural bias. Doesn't accept AI uncritically.
Foundation · D
teacher.foundation.domainD
Sophisticated evaluation — accuracy, depth, nuance, pedagogical fit. Spots subtle bias. Handles AI integrity concerns without false accusations.
Practitioner · D
teacher.practitioner.domainD
Contributes to school/network evaluation of tools, policies, integration outcomes. Leads ethical discussions. Advises leadership on emerging risks.
Leader · D
teacher.leader.domainD
Domain E
Professional Contribution
Professional Contribution
Engages with school AI initiatives. Shares basic tips with immediate colleagues. Asks for support when uncertain.
Foundation · E
teacher.foundation.domainE
Actively shares effective practices. Contributes to department/school discussions. Supports less experienced colleagues. Participates in PLCs.
Practitioner · E
teacher.practitioner.domainE
Leads PD sessions at school/network level. Contributes to AI policy development. Represents the school externally. Documents and shares effective practices.
Leader · E
teacher.leader.domainE
Evidence & Provenance · Part 1 · Teacher
A teacher framework only works if it cites real authority.
The 5 domains × 3 levels above are not Eduversal's opinion. They are mapped block-by-block onto UNESCO's global teacher framework, then anchored to Cambridge International, peer-reviewed research, and the live working patterns of practising teachers. Skim the wall in 60 seconds; click any card to read the original.
15
Sources cited
across the matrix
across the matrix
15
UNESCO competency
blocks mapped 1-to-1
blocks mapped 1-to-1
61%
Empirical anchor —
differentiation use rate
differentiation use rate
2026
Freshness anchor
(latest source)
(latest source)
The three pillars
If you only read three sources, read these. Every cell in the matrix above ultimately points to one of them.
UNESCO
AI Competency Framework for Teachers
First global framework defining teacher AI competencies. 5 aspects × 3 levels = 15 blocks. Eduversal v1.0 Part 1 maps to it block-by-block — see
manifest.externalAnchors.unescoAiCft.Cambridge International
Teaching, Learning, Assessment & AI Hub
The curriculum + coursework authority for our partner schools. Defines the boundary between "AI assists" and "AI authors" in assessment — which the teacher framework operationalises.
Anthropic Education Report
How Educators Actually Use Claude
Real usage data from practising teachers. Differentiation = ~61% of educator use. The empirical reason "differentiate with AI" sits at Practitioner level instead of being aspirational.
Cambridge International
Curriculum + assessment authority
-
The Cambridge Approach to Generative AI & Assessment
Position statement. Anchors Domain B "Ethics" and the Practitioner-level AI-use-in-feedback rows.
-
Cambridge Drives a New Era of Digital Maturity
Feb 2026 Primary + Lower Secondary Digital Literacy refresh — the curriculum signal that "AI literacy is now a digital-maturity competency, not a tech-elective."
UNESCO + United Nations
Global ethics + teacher framework foundations
-
Recommendation on the Ethics of AI
UNESCO's foundational ethics document. Grounds Domain B "Ethics" and the cross-page Red Lines.
-
Guidance for Generative AI in Education & Research
Operational generative-AI guidance cited as a primary source by the AI CFT. Anchors Domain D "Pedagogy" practitioner-level rows.
Indonesia regulation
National data-protection + curriculum context
-
UU PDP No. 27/2022 — Personal Data Protection Law
Indonesia's GDPR-equivalent. Enforced Oct 2024; Lembaga PDP operational 2026. Why the "no student PII in free-tier AI" red line is law, not policy.
-
Permendikdasmen 13/2025 — Koding & AI elective
National regulation. Governs WHAT students learn ABOUT AI; this framework governs HOW teachers USE AI. Sister docs, not overlap.
AI provider educator-tier
The "core three" most likely network-approved
-
Claude for Education (Anthropic)
Learning Mode — Socratic guiding, not direct answering. Strong Bahasa Indonesia per 2025-2026 benchmark. Preferred for HQ curriculum work.
-
ChatGPT for Teachers (OpenAI)
School-tier product, Nov 2025 launch. SAML SSO. Not trained on educator data by default. FERPA-aware. Fastest learning curve for first-time AI users.
-
Gemini for Education (Google)
Bundled with Workspace for Education. Operates as 'school official' under FERPA. Strongest Bahasa Indonesia per 2025-2026 Artificial Analysis Multilingual Index.
Research + standards bodies
Peer-reviewed evidence + assessment integrity
-
What Should Teachers Teach & Students Learn (OECD 2025)
Headline: human-in-the-loop assessment. The "Practitioner can use AI for feedback drafts but never auto-grade" position lives here.
-
GPT detectors are biased against non-native English writers
7 GPT detectors flagged 61% of non-native English essays as AI. Foundational evidence for why Eduversal does NOT recommend AI detection for EAL student work.
Practitioner patterns
Working patterns from real teachers
-
7 AI Prompts That Instantly Upgrade Teaching
Articulates the 4-part mega-prompt pattern (Role + Context + Task + Constraints). The structure behind every prompt in the Eduversal Prompt Library.
-
Simplifying PBL Planning with AI
Iterative prompting beats one-shot for project design. Anchor for Domain D "Pedagogy" Practitioner-level unit-overview workflow.