A global verification layer for human capability and learning
The AICE Framework
A neutral, AI-driven framework for evaluating learning on merit — making skills transparent, comparable, and trusted worldwide.
This is infrastructure that will still make sense in 100 years.
For Governments
Transparent infrastructure for workforce development, skills visibility, and evidence-based policy.
- National skills mapping
- Policy-ready analytics
- Zero regulatory burden
For Organisations
Create verified courses, certify skills, and build trust with AICE-aligned credentials.
- AI-standardised titles
- Complexity scoring
- Public microsite
For Learners
Get your skills recognised on merit, build portable credentials, and prove what you know.
- AICE Points portfolio
- Knowledge Fusion Score
- Verifiable certificates
10 Knowledge Domains
Every course and assessment is tagged with exactly one primary domain and up to two secondary domains, enabling cross-disciplinary learning recognition.
Knowledge Fusion Score (KFS)
When learning spans multiple domains, AICE calculates a Knowledge Fusion Score — recognising the value of cross-disciplinary expertise. A learner studying "Mathematics + Computing" or "Language + Social Sciences" builds verifiable fusion events that demonstrate interconnected understanding.
5 Complexity Bands
All learning content is evaluated and assigned to one of five complexity bands, ensuring fair and consistent comparison across sources.
Introductory
0.25-1.0 AICE PointsFoundational awareness & recall
Foundational
0.5-2.0 AICE PointsCore understanding & concepts
Applied
1.0-4.0 AICE PointsPractical application & skills
Advanced
2.0-8.0 AICE PointsSynthesis & problem-solving
Expert
4.0-15.0 AICE PointsEvaluation & mastery-level judgment
Understanding AICE Complexity Bands
Detailed mappings, distinctions, and cross-system equivalencies for the AICE framework.
How AICE Points Are Calculated
Transparent, auditable, and academically grounded. Every number has a derivation.
AICE Points
A standardised unit of learning effort and complexity. AICE Points are calculated by AI based on content depth, time investment, and assessment rigour.
AI-Calculated
Points are assigned by analysing learning materials, not self-reported. The AI evaluates cognitive load, prerequisite knowledge, and practical application depth.
Comparable
10 AICE Points from any source represents the same learning effort. A course from a university and a practical workshop are measured equally.
Stackable
Points accumulate over time across domains. Your learning portfolio grows continuously, reflecting lifelong skill development.
Title Guardrails
AI-enforced standards ensure all titles are honest, comparable, and free from misleading claims.
Strictly Banned
- Accredited, Certified, Licensed
- Degree, Bachelor, Master, PhD
- Board Certified, Chartered
- Guaranteed, Job Ready
- Government Approved
Auto-Normalised
- MasteryAdvanced
- MasterclassAdvanced
- EliteAdvanced
- Professional LevelAdvanced
- Industry LeadingAdvanced
Canonical titles describe learning scope and level — never authority, permission, or outcomes.
Why AICE Exists
Across the world, governments and organisations face the same structural challenges:
- Skills are hard to measure outside formal degrees
- Learning happens everywhere — but is not comparable or verifiable
- CVs and certificates are inflated, fragmented, or unverifiable
- Workforce upskilling is expensive, slow, and opaque
- Traditional education systems cannot adapt quickly enough
As a result, talent is underutilised, reskilling is inefficient, and opportunity is unevenly distributed.
Built on Transparency
Merit-Based Evaluation
Learning content is analysed using AI to assess complexity, learning effort, and skill depth. No judgement is made on the background of the creator or learner.
Democratic & Inclusive
Anyone can contribute learning material. A skilled mechanic, nurse, or technician can share knowledge just as a university can.
Continuous, Not Static
AICE recognises that learning is ongoing. Credits are modular, stackable, and updateable over time.
Explainable AI
All evaluations are transparent, auditable, and adjustable. AI assists measurement; it does not make opaque decisions.
What AICE Is Not
- AICE does not issue degrees
- AICE does not license professions
- AICE does not replace national education authorities
- AICE does not make hiring or immigration decisions
AICE is measurement infrastructure, not a regulator.
Verification Roadmap
AICE is designed as a progressive verification system.
Our current focus is on transparent, independently verifiable records of professional learning and progression.
Public verification pages allow employers and institutions to confirm the authenticity of major professional milestones, while clear eligibility rules ensure ledger verification is applied proportionately.
Identity verification may be introduced selectively in the future where appropriate, but is not required for general participation.
AICE prioritises verification of professional outcomes. Identity verification is applied selectively where proportionate and necessary.
Contextual AI Assistant
Every public course page includes an AI assistant that answers questions about the certificate or course using stored AICE evaluation data.
What you can ask:
- How were the AICE Points calculated?
- What complexity band is this course?
- What domains does this cover?
- How long is the estimated learning time?
- What assessments are included?
Built-in safeguards:
- No comparisons to other learners
- No endorsements or recommendations
- No qualification claims
- Grounded in stored metadata only
- Transparent reasoning always shown
Pilot Programme Model
We invite governments to participate in observed pilot programmes — not accreditation.
A pilot may include:
- A defined group (workforce, students, trainees)
- Use of AICE to evaluate learning material
- Aggregated, anonymised insights
- No regulatory commitment required
Outcomes:
- Data-driven insight into skills
- Improved learning transparency
- Evidence for future policy decisions
- Skills gap visibility
Low risk
No legal changes
High visibility
Clear metrics
Cost-efficient
Digital scale
Inclusive
All learning
Future-ready
AI-native
Phased Verification Approach
AICE follows a phased and proportionate approach to verification.
Initial implementation focuses on public verification pages that allow employers and institutions to independently confirm the authenticity of professional milestones. Ledger attestation is applied selectively to significant achievements, ensuring high-signal verification without unnecessary complexity.
Eligibility rules are configurable by administrators, ensuring governance flexibility. Identity verification is not required for general participation and may only be introduced for high-stakes use cases where proportionate.
AICE applies verification where it adds trust, not friction.
View Full Execution PlanGovernance & Ethics
- Transparency by design
All AI decisions are explainable and auditable
- Clear separation of roles
AICE measures; it does not regulate or endorse
- User consent and control
Learners own their data and credentials
- No hidden algorithms
Scoring logic is documented and reviewable
Centralised Intelligence, Decentralised Execution
The governing philosophy of AICE's AI architecture.
Luna — Central Orchestrator
AICE's master orchestration agent
Luna routes content through the scoring pipeline, coordinates specialist agents, and manages the audit trail — but executes no irreversible actions herself.
Intelligence is centralised in Luna's orchestration layer. Execution is decentralised across deterministic systems and human-approved workflows. This separation ensures that AI enhances decision-making without becoming an unchecked authority.
Temperature 0 Scoring
All classification, extraction, and scoring agents run at temperature 0 for deterministic, reproducible results across identical inputs.
Versioned System Prompts
Every agent's system prompt is version-controlled in the database. Changes require governance approval and create an immutable audit trail.
No Direct Ledger Writes
AI agents can recommend, classify, and score — but cannot mint certificates, write to the verification ledger, or execute irreversible actions without human or deterministic system authorisation.
Explainable Outputs
A dedicated explainability agent (temperature 0.3) translates technical scores into human-readable rationale, ensuring every decision can be understood and challenged.
"AI advises. Humans and deterministic systems execute."
A World Where Learning Speaks for Itself
We envision a world where skills are recognised on merit, regardless of where or how they were learned. Where a self-taught developer, a vocational nurse, and a university graduate are measured on the same transparent scale.
International Standards Framework
The International AI Credit System
(IACS) Framework
1 AICE Point = 1 Notional Learning Hour
This is the foundational unit of the IACS framework. Every certificate, qualification, and professional credential in the system is built on this single atomic unit — a verified hour of level-appropriate learning.
Qualification Pathway
The 7-Level Framework
Credit Accumulation
The Credit Stack
AICE Points accumulate into recognised qualification tiers
Ready to get started?
Whether you're a government exploring pilots, an organisation creating courses, or a learner building your portfolio — AICE is ready.