AI Collaboration Framework
A structured approach that helps people collaborate with AI systematically, critically and responsibly—from understanding the problem to producing a solution.
in the Loop
From a single prompt to a complete process.
Many users give one instruction, receive one answer and then use it immediately.
Real tasks, however, are rarely that simple. Problems need to be understood, context needs to be built, suggestions need to be evaluated and outcomes need to be refined.
The AI Collaboration Framework transforms interaction with AI from a one-off transaction into a human-led collaborative process.
AI helps accelerate and expand the process. Humans set the direction, evaluate meaning and remain accountable.
Principles embedded within the framework.
These principles do not stand apart as separate theory. They guide every step of the collaborative process.
Human First
Purpose, judgement and final decisions always remain with the human.
Context First
The quality of the outcome depends on how well the situation and its requirements are understood.
Think Critically
Every response must be evaluated, reviewed and compared before it is used.
Act Responsibly
Accuracy, ethics, privacy and the consequences of use must never be overlooked.
Nine components. One flow of thought.
Each component has its own function, yet all are interconnected. Humans remain actively involved from beginning to end.
Problem
Identify what truly needs to be understood, solved or produced.
Conversation
Use conversation to explain the task, audience, situation and constraints.
Context
Add information, documents, examples and constraints progressively.
Critical Thinking
Examine assumptions, gaps, bias, logic and the suitability of the suggestions provided.
Prompt Chaining
Break complex tasks into smaller, connected steps.
Evidence
Check the facts, sources, data and supporting documents being used.
Review
Evaluate accuracy, tone, suitability, ethics and the implications of use.
Refinement
Refine the outcome through corrections, clarification and follow-up instructions.
Final Output
The final output is the outcome that has been selected, reviewed and improved.
Humans are not only present at the end of the process.
In AI Collaboration, humans remain involved throughout the entire cycle—setting direction, building context, evaluating suggestions, reviewing evidence, refining outcomes and making the final decision.
in the Loop
Not a straight line. It is a cycle.
Good collaboration moves forward, but also returns to earlier stages when new information or weaknesses are identified.
Real processes are always in motion.
A review may reveal that the context is still insufficient. New evidence may change the decision. Feedback may require the conversation to begin again.
Examples of movement within the cycle
A framework becomes meaningful when it is applied.
See how AI Collaboration is applied in writing, education, research, administration and real organisational work.
Explore AI Collaboration in Action →