CloudPedagogy Frameworks

Complementary frameworks for developing Human–AI capability, designing augmented cognition, and building responsible AI-enabled systems.

Artificial intelligence increasingly participates not only in tasks, but in learning, reasoning, professional judgement, research, decision-making, and organisational systems.

CloudPedagogy provides three complementary frameworks and methods addressing different dimensions of Human–AI practice:

CAPABILITY → COGNITION → SYSTEM DESIGN

Each addresses a distinct design question. They can be used independently or together depending on the purpose and context.

1. AI Capability Framework

CAPABILITY

What capabilities do people and institutions need to work responsibly and effectively with AI?

The CloudPedagogy AI Capability Framework (AICF) is a six-domain developmental framework for building durable Human–AI capability across professional and institutional contexts.

It moves beyond tool proficiency towards the capabilities required to understand AI, work with it intentionally, exercise judgement, recognise ethical and societal implications, govern AI-influenced practice, and continue learning as technologies and contexts change.

Why Capability Matters

AI skill is about using tools.
AI capability is about exercising judgement.

A person may be technically proficient with AI while still lacking the ability to determine when it should be used, how its contributions should be evaluated, what responsibilities remain human, what consequences may follow, or when AI recommendations should be challenged or rejected.

Capability therefore extends beyond AI literacy or tool proficiency. It encompasses the ability to work effectively with AI while preserving human judgement, ethical awareness, accountability, and the capacity to adapt as technologies and professional contexts evolve.

The Six Domains

The Framework is organised into six interdependent domains:

AI Awareness & Orientation
Understanding AI systems, their possibilities, limitations, uncertainty, and context.

Human–AI Co-Agency
Establishing intentional and accountable relationships between human and machine intelligence.

Applied Practice & Innovation
Applying AI purposefully while exploring new forms of practice and innovation.

Ethics, Equity & Impact
Recognising ethical implications, anticipating harms and benefits, and considering impacts on people and communities.

Decision-Making & Governance
Exercising judgement, oversight, and accountability where AI influences decisions and professional practice.

Reflection, Learning & Renewal
Learning from Human–AI practice and adapting capability as technologies, contexts, and responsibilities change.

Together, the six domains provide a developmental model for responsible and adaptive Human–AI capability.

Core question:
What Human–AI capability needs to be developed?

📥 AI Capability Framework on Zenodo

Version: 2026 Edition
Licence: Creative Commons Attribution–NonCommercial–ShareAlike 4.0 International (CC BY-NC-SA 4.0)

2. Human–AI Augmented Cognition Framework

COGNITION

What cognition matters — and where should it occur?

The Human–AI Augmented Cognition Framework (HAACF) is a purpose-led framework for designing and evaluating how cognition is distributed within Human–AI activity.

Rather than beginning with what an AI system can perform, HAACF begins by identifying the cognition that has value within an activity and why that cognition matters.

It then provides a structured method for considering whether particular cognitive operations should be:

RETAINED
Substantially within the human component where human performance of the cognition continues to matter.

AUGMENTED
Through complementary Human–AI interaction where meaningful human cognitive participation should be preserved or strengthened.

DELEGATED
Substantially to artificial intelligence where human performance provides insufficient continuing value and appropriate assurance exists.

Where an adequately justified allocation cannot yet be established, HAACF also recognises an UNRESOLVED decision state.

The Cognitive Allocation Test

HAACF considers cognitive allocation through seven dimensions:

Capability • Responsibility • Consequence • Context • Evaluability • Complementarity • Efficiency

Designing Augmentation

Where augmentation is appropriate, HAACF provides a five-part augmentation repertoire for structuring Human–AI cognitive interaction:

FORM • EXPAND • CHALLENGE • EVALUATE • JUDGE

HAACF also distinguishes between Task Performance and Human Capability, recognising that improved immediate performance does not necessarily mean that important human capability has been strengthened or preserved.

The Framework incorporates Evaluability, Evaluation Escalation, Metacognitive Regulation, and Reflect & Adapt so that cognitive allocation remains deliberate, examinable, and revisable.

Core question:
What cognition matters — and where should it happen?

📥 Human–AI Augmented Cognition Framework v1.0 on Zenodo

Version: 1.0
Licence: Creative Commons Attribution 4.0 International (CC BY 4.0)

3. Capability-Driven Development

SYSTEM DESIGN

How should capability and Human–AI requirements shape the systems we build?

Capability-Driven Development (CDD) is a capability-first approach to designing AI-enabled and data-informed systems.

Rather than beginning with what technology can automate, CDD begins with the human and institutional capabilities that a system should support, preserve, or strengthen.

CDD translates capability, responsibility, governance, and Human–AI boundary requirements into explicit system-design decisions and artefacts.

This includes:

defining appropriate Human–AI boundaries within workflows
identifying capabilities that systems should preserve or strengthen
embedding transparency, oversight, and accountability
designing processes that remain inspectable and reviewable
supporting escalation, override, and responsibility traceability
evaluating systems throughout their lifecycle

Practical Design Artefacts

CDD supports system design through practical artefacts including:

System Capability Briefs
Human–AI Boundary Maps
Risk and Misuse Registers
Governance and Oversight Plans
Evaluation Logs
Retirement Notes

CDD complements rather than replaces established engineering and development practices such as Agile and DevOps.

Core question:
How should capability requirements and Human–AI boundaries be translated into system design?

📥 Capability-Driven Development on Zenodo

💻 Capability-Driven Development on GitHub

How the Three Frameworks Relate

The three approaches address different levels of the broader Human–AI design problem.

Framework	Primary Concern	Core Question
AI Capability Framework	Capability	What capabilities do people and institutions need?
Human–AI Augmented Cognition Framework	Cognition	What cognition matters, and where should it occur?
Capability-Driven Development	System Design	How should capability requirements and Human–AI boundaries shape system design?

A simple way to understand their relationship is:

CAPABILITY
AI Capability Framework
↓

COGNITION
Human–AI Augmented Cognition Framework
↓

SYSTEM DESIGN
Capability-Driven Development

This does not mean that every activity must use all three approaches or that they must always be applied sequentially.

Each framework can be used independently. Where they are combined, however, they provide complementary perspectives from human capability, through cognitive allocation, to system design and implementation.

An Example of the Frameworks Working Together

Consider the design of an AI-supported research appraisal process.

AI Capability Framework — CAPABILITY

The organisation identifies the capabilities researchers require: AI awareness, evidence evaluation, Human–AI Co-Agency, ethical judgement, governance awareness, and reflective practice.

↓

HAACF — COGNITION

The research activity is decomposed into cognitive operations. Routine evidence retrieval might be Delegated, generation of alternative interpretations might be Augmented, and consequential evidential evaluation or judgement might remain substantially human.

↓

CDD — SYSTEM DESIGN

Those requirements become explicit system-design decisions: AI roles are defined, human review points are established, consequential actions require appropriate authorisation, evaluation mechanisms are provided, and responsibilities remain traceable.

The frameworks therefore operate at different levels without needing to be collapsed into a single model.

📦 AI Capability Framework Resources

The AI Capability Framework is designed for operational use in teaching, research, professional practice, facilitation, and institutional development.

Explore the AI Capability Framework resources on GitHub

🎓 Free AI Capability Framework Orientation Course

The AI Capability Framework Orientation Course provides a structured introduction to the six domains.

Participants explore the Framework through applied scenarios and reflection and develop a personal capability action plan.

Free for all users.

Access the AI Capability Framework Orientation Course

📚 Framework-Based Courses

CloudPedagogy courses apply Human–AI capability development to a range of professional and societal contexts, including:

Equity, Inclusion & Social Justice
Policy & Governance
Creative & Cultural Practice
Global & Public Health
Humanitarian Crisis Response
Climate Change & Sustainability

Explore Framework-Based Courses

🔗 Citing the Frameworks

For formal citation, use the citation information provided by the relevant Zenodo record:

📜 Licence and Use

The individual CloudPedagogy frameworks are published under the licence specified in their respective publication and Zenodo records.

The AI Capability Framework is licensed under CC BY-NC-SA 4.0.

The Human–AI Augmented Cognition Framework v1.0 is licensed under CC BY 4.0.

For commercial licensing enquiries relating to materials where commercial use is restricted, institutional adoption, or consultancy enquiries:
[email protected]