Level 1
Unconscious Incompetence
Discover & Awaken
How it’s confirmed
Self-report only
Capability reference
AI Fluency Framework
25 skills. Five domains. Five levels.
25 skills across 5 domains, Level 1 to Level 5.
How the framework works
Maturity is a claim until there’s evidence.
The AI Fluency Framework uses one five-level ladder across every skill. The ladder starts with first awareness, moves through deliberate and repeatable practice, and ends with work that other people can use.
A claimed level and a confirmed level aren’t the same thing. Levels 1 and 2 can begin with self-report, but the standard rises with the claim: work samples at Level 3, observation at Level 4, and evidence that someone else adopted what you built at Level 5.
Level 2
Conscious Incompetence
Explore & Understand
How it’s confirmed
Self-report + named gap
Level 3
Conscious Competence
Apply & Practise
How it’s confirmed
Work-sample review
Level 4
Unconscious Competence
Intuit & Adapt
How it’s confirmed
Peer / manager observation
Level 5
Mastery
Create & Transform
How it’s confirmed
Adoption by others
Domain A
A · AI Fundamentals
Concepts, privacy, bias and limitations — the working knowledge that makes every other skill safe to apply.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Treats AI as a search engine retrieving pre-existing correct answers rather than a system generating text token by token. Cannot explain why a confident response might be wrong.
How it’s confirmed
Describes AI as “looking things up” when asked directly.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Understands AI generates rather than retrieves, but cannot explain training data, tokens, context windows, or why hallucination happens.
How it’s confirmed
Can say “I don’t really get why it makes things up” but not why.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Explains training data, prediction, tokens, context and hallucination in practical terms suited to a colleague, and sets realistic expectations before starting a task.
How it’s confirmed
Gives a written or verbal explanation of a recent AI mistake that correctly names the mechanism behind it.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Adapts the explanation instinctively to the audience and spots a colleague’s misconception about AI before it causes a problem.
How it’s confirmed
A colleague notes they corrected a misunderstanding about AI in a meeting without being asked.
Level 5
Mastery
Create & Transform
What it looks like
Builds frameworks, training material or onboarding content that teaches AI concepts to others at scale.
How it’s confirmed
An onboarding doc or explainer they wrote is used by people outside their own team.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Assumes AI can complete most knowledge-based tasks whenever the right tool exists, without weighing the task itself.
How it’s confirmed
Says “I just try AI on most things.”
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Senses that some tasks suit AI better than others but can’t articulate the distinguishing factors — accuracy needs, available information, cost of error.
How it’s confirmed
Can identify a task AI failed at but not predict which task will fail next.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Deliberately weighs the task, the information available, the accuracy required and the cost of being wrong before deciding to use AI.
How it’s confirmed
A documented decision, even informal, showing the task was assessed before AI was applied.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Identifies instantly where AI adds value, where human judgement is required, and which hybrid approach fits — without stopping to reason it through.
How it’s confirmed
Observed making a fast, correct AI/no-AI call under time pressure.
Level 5
Mastery
Create & Transform
What it looks like
Designs decision frameworks — checklists, criteria, triage rules — that help a team systematically identify which tasks suit AI.
How it’s confirmed
A triage framework or rule-of-thumb they created is used by the team to make AI/no-AI calls.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Enters whatever information the task requires unless a system actively blocks it, without considering sensitivity.
How it’s confirmed
Describes pasting client or personal data into AI tools without a second thought.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Knows sensitive data creates risk but is often unsure what specifically is safe to enter, so decisions are inconsistent.
How it’s confirmed
Hesitates before entering data but can’t articulate the rule they’re applying.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Checks the sensitivity of the information, the tool’s approved use, and relevant policy before entering anything into an AI system.
How it’s confirmed
Can point to a specific instance where data was withheld or redacted based on a stated policy.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Instinctively anonymises, redacts or restructures sensitive data before it goes near an AI tool, and flags the risk to others unprompted.
How it’s confirmed
A colleague reports being warned about a data-handling risk they hadn’t noticed.
Level 5
Mastery
Create & Transform
What it looks like
Helps design the data-handling standards, tool approvals or technical controls that let an organisation use AI safely at scale.
How it’s confirmed
A data-handling guideline or control they authored is part of official policy.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Assumes an AI recommendation about people is objective because it came from technology rather than a human.
How it’s confirmed
Accepts AI-generated people-decisions (hiring, scoring, ranking) at face value.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Knows AI can be biased but isn’t sure how to detect it in a specific output or what to do once suspected.
How it’s confirmed
Flags “this might be biased” without being able to say why or how to check.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Examines the inputs, assumptions and affected groups behind a recommendation before relying on it, and can name a specific concern.
How it’s confirmed
A written note or comment showing bias was actively checked on a real output.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Anticipates likely sources of bias before they show up and adjusts the prompt, data or review process pre-emptively.
How it’s confirmed
Observed changing an approach specifically to reduce a bias risk before anyone raised it.
Level 5
Mastery
Create & Transform
What it looks like
Creates evaluation methods, safeguards or policies that systematically reduce unfair or discriminatory AI outcomes.
How it’s confirmed
A bias-check process or safeguard they built is in standard use.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Trusts a confident, well-written AI response unless something looks obviously wrong on the surface.
How it’s confirmed
Describes rarely double-checking AI outputs before using them.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Knows a response might be wrong but doesn’t consistently know which specific parts need verification.
How it’s confirmed
Checks “sometimes,” inconsistently, with no rule for when.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Verifies claims, assumptions, calculations and sources based on how much the task’s risk level demands.
How it’s confirmed
A real example where a specific claim was fact-checked and corrected before use.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Recognises the patterns that signal fabrication or overconfidence — vague citations, suspiciously round numbers — and adapts on the spot.
How it’s confirmed
Caught a fabricated detail that others missed, without a formal review step.
Level 5
Mastery
Create & Transform
What it looks like
Designs verification standards or QA processes that make reliable AI-assisted work repeatable across a team.
How it’s confirmed
A verification checklist or QA step they created is required practice for others.
Domain B
B · AI Interaction
Prompting, context and evaluation — the moment-to-moment craft of getting a good result out of a model.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Enters a short request and expects the system to infer everything else about what’s needed.
How it’s confirmed
Describes typing one-line requests as the default.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Adds more detail only reactively, after the first response turns out to be unhelpful.
How it’s confirmed
Knows first prompts are usually too thin but doesn’t front-load detail.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Deliberately states the objective, context, audience, constraints and required output format before sending the first prompt.
How it’s confirmed
A saved prompt for a real task that explicitly states all four elements.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Structures prompts instinctively and adapts technique automatically depending on the tool or model being used.
How it’s confirmed
Observed switching prompting style unprompted when moving between tools.
Level 5
Mastery
Create & Transform
What it looks like
Creates reusable prompting methods, templates or interaction systems that other people rely on.
How it’s confirmed
A prompt template they built is actively used by someone else on the team.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Concludes the tool isn’t capable of the task after one weak response, rather than adjusting the approach.
How it’s confirmed
Describes abandoning AI after a single bad result.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Rewords the question or asks the model to try again, without diagnosing what actually went wrong.
How it’s confirmed
Retries repeatedly with only surface-level changes.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Diagnoses what’s specifically missing from a weak response and gives targeted feedback, examples or constraints to fix it.
How it’s confirmed
A real before/after showing a specific, named fix that improved the output.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Fluidly guides the model through several refinement steps, changing technique based on how it responds each time.
How it’s confirmed
Observed running a multi-step refinement live without pausing to plan it.
Level 5
Mastery
Create & Transform
What it looks like
Designs structured interaction processes — multi-step prompting patterns — that reliably produce strong outcomes for others.
How it’s confirmed
A refinement process they documented is followed by colleagues on similar tasks.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Provides minimal context on the assumption that the AI should already understand the topic.
How it’s confirmed
Describes giving little to no background before asking for output.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Knows context matters but tends to over-share unstructured material or under-share relevant detail, inconsistently.
How it’s confirmed
Pastes whole documents “just in case” without curating them.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Selects and organises the information most relevant to the task and the desired output before providing it.
How it’s confirmed
A real example showing context was deliberately trimmed or structured before sending.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Instinctively adjusts the amount, sequence and format of context to improve results, without conscious planning.
How it’s confirmed
Observed re-sequencing context mid-task and getting a visibly better result.
Level 5
Mastery
Create & Transform
What it looks like
Designs context architectures, knowledge structures or retrieval approaches for AI systems used by others.
How it’s confirmed
A context template or knowledge structure they built is reused across multiple tasks or people.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Judges a response mainly by whether it sounds clear and convincing on first read.
How it’s confirmed
Describes accepting fluent-sounding output without further checks.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Checks a response informally against what they already know, without a consistent method.
How it’s confirmed
Sometimes catches errors, sometimes doesn’t, with no repeatable process.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Assesses reasoning, evidence, assumptions and output against the specific requirements of the task.
How it’s confirmed
A real example where a response was checked against a stated requirement and found lacking.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Evaluates a response from multiple angles quickly and knows when it warrants deeper testing.
How it’s confirmed
Observed flagging a subtle issue in an AI output that a less experienced colleague missed.
Level 5
Mastery
Create & Transform
What it looks like
Develops evaluation criteria, benchmarks or QA methods for AI-generated work more broadly.
How it’s confirmed
An evaluation rubric or checklist they created is used to assess others’ AI output.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Accepts a recommendation as long as the explanation given sounds logical on its face.
How it’s confirmed
Describes taking AI’s stated reasoning at face value.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Occasionally asks the AI whether it’s certain, or requests another answer, without a structured challenge.
How it’s confirmed
Pushes back sometimes but not systematically.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Asks for assumptions, alternatives, counterarguments, evidence and potential failure points before accepting a recommendation.
How it’s confirmed
A real transcript or note showing the recommendation was actively stress-tested.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Uses multiple testing techniques and compares tools, models or independent sources without needing a fixed process.
How it’s confirmed
Observed cross-checking an AI recommendation against an independent source unprompted.
Level 5
Mastery
Create & Transform
What it looks like
Designs adversarial testing or evaluation systems that surface weaknesses in AI outputs before they reach a high-stakes decision.
How it’s confirmed
A stress-test process or checklist they created is used before high-stakes AI-assisted decisions.
Domain C
C · AI Application
Workflows, automation and measurement — turning individual skill into work that actually gets redesigned around AI.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Uses AI rarely, and mostly out of curiosity rather than as part of how work gets done.
How it’s confirmed
Describes infrequent, exploratory use.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Experiments with AI regularly but it hasn’t become part of a consistent working method yet.
How it’s confirmed
Uses AI often but inconsistently, task by task.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Uses AI deliberately for defined tasks and produces reliable results from that use.
How it’s confirmed
A named, repeatable task where AI is consistently part of the workflow.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
AI is naturally integrated into how they work, and its role adapts automatically to the situation.
How it’s confirmed
A colleague notes AI use as simply part of how this person works, not a separate activity.
Level 5
Mastery
Create & Transform
What it looks like
Creates new AI-enabled working models that materially change how a piece of work gets performed.
How it’s confirmed
A new way of working they introduced has been adopted by others on the team.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Uses AI as a bolt-on activity that sits alongside the existing process without changing it.
How it’s confirmed
Describes using AI for one step while the rest of the process stays the same.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Can see opportunities to integrate AI into a process but isn’t sure exactly where it should fit.
How it’s confirmed
Identifies the opportunity but hasn’t mapped where AI belongs.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Maps the process, identifies the right AI touchpoints, and defines where human review is still required.
How it’s confirmed
A documented process map showing AI touchpoints and review gates.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Fluidly redesigns workflows around the respective strengths of people, AI and other systems.
How it’s confirmed
Observed redesigning a workflow on the fly as new AI capability became available.
Level 5
Mastery
Create & Transform
What it looks like
Designs scalable operating models in which AI transforms the end-to-end workflow, not just a step within it.
How it’s confirmed
An operating model or redesigned workflow they built is running across a team or function.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Continues doing repetitive work manually unless someone else provides an automated solution.
How it’s confirmed
Describes waiting for others to build automation rather than attempting it.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Recognises an automation opportunity but needs help selecting or configuring the tools to act on it.
How it’s confirmed
Names the opportunity but can’t yet build the solution unassisted.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Builds or configures a basic automation and tests whether it works reliably before relying on it.
How it’s confirmed
A working automation they built, with evidence it was tested.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Routinely connects tools, data and AI components to automate work, managing exceptions as they come up.
How it’s confirmed
Observed handling an automation exception live without escalating it.
Level 5
Mastery
Create & Transform
What it looks like
Architects automation systems that can be governed, maintained and scaled across multiple teams or processes.
How it’s confirmed
An automation system they built is in active use beyond their own workflow.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Asks AI for the answer directly and uses what it provides without further exploration.
How it’s confirmed
Describes taking a single AI answer on a complex or unfamiliar problem.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Uses AI to explore a topic but can get overwhelmed by conflicting or uncertain information.
How it’s confirmed
Explores broadly but struggles to converge on an answer.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Breaks the problem into parts, tests assumptions, and combines AI assistance with relevant human expertise.
How it’s confirmed
A real example showing the problem was decomposed and assumptions were checked.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Moves fluidly between exploration, analysis, experimentation and judgement as new information emerges.
How it’s confirmed
Observed adapting their approach mid-problem as new information changed the picture.
Level 5
Mastery
Create & Transform
What it looks like
Creates new problem-solving methods that combine AI, data, systems and human expertise in ways not done before.
How it’s confirmed
A problem-solving method they developed has been used by others on a different problem.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Judges success mainly by whether the AI output appears useful at a glance.
How it’s confirmed
Describes no formal measurement of whether AI use is paying off.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Considers whether AI saved time but doesn’t consistently measure the result.
How it’s confirmed
Has a rough sense of value but no numbers behind it.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Compares outcomes — quality, time, cost, risk or user experience — against the previous, non-AI approach.
How it’s confirmed
A real before/after comparison with named metrics.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Continuously refines an AI application using performance data, feedback and changing requirements.
How it’s confirmed
Observed adjusting an AI-based process specifically because the numbers said to.
Level 5
Mastery
Create & Transform
What it looks like
Designs measurement systems that demonstrate and optimise AI’s strategic and operational value more broadly.
How it’s confirmed
A measurement framework or dashboard they built is used to justify or steer AI investment.
Domain D
D · AI Strategy
Priorities, risk, governance and scaling — the judgement that connects AI decisions to the business, not just the task.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Pursues AI initiatives because the technology is interesting or impressive, not because of a clear business need.
How it’s confirmed
Describes chasing AI ideas for their novelty.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Knows an initiative should support a business need but finds it difficult to connect the technology to a measurable priority.
How it’s confirmed
Can name the technology’s appeal but not the business case.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Defines the business problem, intended outcome, users and measure of success before proceeding with an AI initiative.
How it’s confirmed
A written problem statement and success measure for a real initiative.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Instinctively connects AI opportunities to strategic priorities, operating constraints and stakeholder needs.
How it’s confirmed
Observed reframing a proposed AI idea in terms of strategic priority, unprompted.
Level 5
Mastery
Create & Transform
What it looks like
Shapes organisational strategy by identifying how AI can create genuinely new sources of value or advantage.
How it’s confirmed
A strategic direction they proposed has been adopted at an organisational level.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Favours whichever AI idea uses the newest or most powerful technology, regardless of fit.
How it’s confirmed
Describes choosing initiatives by technical novelty.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Considers potential value when choosing between AI ideas, but without a consistent prioritisation method.
How it’s confirmed
Has instincts about value but no repeatable comparison method.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Compares opportunities using value, feasibility, risk, cost, readiness and learning potential as explicit criteria.
How it’s confirmed
A written comparison of two or more real opportunities against named criteria.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Balances strategic and practical considerations rapidly and adjusts priorities as new evidence arrives.
How it’s confirmed
Observed re-ranking priorities live in a meeting as new information came in.
Level 5
Mastery
Create & Transform
What it looks like
Creates portfolio approaches that sequence AI investments and capability-building across an organisation.
How it’s confirmed
A prioritisation framework or portfolio view they built is used to plan AI investment.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Considers AI risk only after a problem has already occurred or someone has raised a concern.
How it’s confirmed
Describes reacting to risk after the fact, not before.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Knows risk should be considered but is unsure which risks are actually relevant to a given initiative.
How it’s confirmed
Flags “there might be risk” without naming which kind.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Identifies risks — privacy, accuracy, security, bias, intellectual property, human impact — before deployment.
How it’s confirmed
A written risk assessment covering named categories for a real initiative.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Naturally incorporates proportionate controls throughout design, testing and operation, without a separate risk step.
How it’s confirmed
Observed building a control into a design before anyone asked for a risk review.
Level 5
Mastery
Create & Transform
What it looks like
Designs integrated risk frameworks that enable innovation while protecting the organisation and affected stakeholders.
How it’s confirmed
A risk framework they built is applied to AI initiatives beyond their own.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Treats the AI system or its provider as primarily responsible for a flawed outcome.
How it’s confirmed
Attributes a bad outcome to “the AI” rather than a decision-owner.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Understands a person should remain accountable but the boundaries of that accountability aren’t always clear to them.
How it’s confirmed
Agrees someone should own the decision but can’t specify who or where.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Defines decision ownership, review requirements, escalation points and acceptable-use boundaries for a given use case.
How it’s confirmed
A written accountability map or acceptable-use statement for a real initiative.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Adapts governance to the specific context while keeping accountability and oversight consistently clear.
How it’s confirmed
Observed clarifying an ownership question in real time without needing a policy lookup.
Level 5
Mastery
Create & Transform
What it looks like
Designs governance models that support responsible AI use across different functions, risk levels and degrees of autonomy.
How it’s confirmed
A governance model they designed is applied across multiple teams or use cases.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Keeps using a successful AI approach personally and assumes others will pick it up if it’s useful.
How it’s confirmed
Describes no active effort to spread a working approach.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Shows others what they’ve done, but adoption beyond that stays informal and inconsistent.
How it’s confirmed
Has shared the idea but can’t point to sustained uptake.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Documents the use case, process, controls and expected benefit so others can reproduce it themselves.
How it’s confirmed
A real document that lets someone else reproduce the approach without help.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Adapts the solution for different users, coaches others directly, and actively addresses barriers to sustained adoption.
How it’s confirmed
Observed troubleshooting someone else’s adoption of the approach.
Level 5
Mastery
Create & Transform
What it looks like
Creates the standards, platforms, capability programs and operating conditions required to scale AI use across an organisation.
How it’s confirmed
A program or platform they built is the reason a use case scaled beyond one team.
Domain E
E · AI Growth
Experimentation, learning and knowledge sharing — the habits that keep fluency compounding instead of going stale.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Waits until someone else demonstrates that a new AI tool or technique is relevant before trying it.
How it’s confirmed
Describes following rather than initiating experimentation.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Explores new tools but often without a clear objective or method behind the exploration.
How it’s confirmed
Tries things but can’t say what they were testing for.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Defines a practical hypothesis or use case, tests it, and assesses the outcome against that hypothesis.
How it’s confirmed
A real experiment with a stated hypothesis and a recorded result.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Runs frequent, focused experiments and quickly incorporates useful discoveries into their actual work.
How it’s confirmed
Observed adopting a new technique into live work within days of testing it.
Level 5
Mastery
Create & Transform
What it looks like
Establishes experimentation systems that enable other people to discover and scale valuable new applications.
How it’s confirmed
An experimentation process or forum they set up is used by others to test new ideas.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Continues using familiar methods until they visibly stop working, rather than tracking change proactively.
How it’s confirmed
Describes learning about new capability only after being forced to.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Reads or watches updates about new AI capability but struggles to translate that into practical skill.
How it’s confirmed
Is aware of a new capability but hasn’t used it yet.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Identifies relevant changes, practises using them, and incorporates useful techniques into their actual work.
How it’s confirmed
A real example of a newly learned capability applied to a task.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Quickly understands the implications of a new capability and adapts their methods without needing formal instruction.
How it’s confirmed
Observed using a newly released capability correctly before formal guidance existed.
Level 5
Mastery
Create & Transform
What it looks like
Anticipates capability shifts ahead of time and helps the organisation prepare for how they’ll change work, products or strategy.
How it’s confirmed
A heads-up or preparation plan they raised proactively shaped how the team responded to a change.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Assumes a tool has become unreliable and stops using a workflow once it produces worse results.
How it’s confirmed
Describes abandoning a workflow rather than diagnosing what changed.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Tries minor changes to fix a broken workflow but often needs someone else to explain what actually changed.
How it’s confirmed
Attempts fixes without understanding the underlying cause.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Retests the workflow, isolates the cause, and deliberately adjusts the prompts, model, data or process.
How it’s confirmed
A real diagnosis showing the cause was isolated and a specific fix applied.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Adapts quickly by drawing on a broad understanding of how different tools and models tend to behave.
How it’s confirmed
Observed fixing a broken workflow fast, without needing to research the cause from scratch.
Level 5
Mastery
Create & Transform
What it looks like
Redesigns systems to stay resilient as the underlying technologies and capabilities continue to evolve.
How it’s confirmed
A resilience pattern or fallback design they built protects a workflow others depend on.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Uses an effective technique themselves but doesn’t generally share it with others.
How it’s confirmed
Describes keeping useful discoveries to themselves.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Mentions a technique to colleagues or forwards the prompt, without further explanation.
How it’s confirmed
Shares the “what” but not the “why” or “how.”
Level 3
Conscious Competence
Apply & Practise
What it looks like
Explains the method, suitable use cases, limitations, and the steps needed for someone else to reproduce the result.
How it’s confirmed
A real write-up that lets someone reproduce the technique from the explanation alone.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Actively coaches others and adapts the explanation to their role, experience level and specific needs.
How it’s confirmed
Observed tailoring an explanation differently for two different colleagues.
Level 5
Mastery
Create & Transform
What it looks like
Creates knowledge-sharing systems, communities or capability programs that improve collective AI fluency.
How it’s confirmed
A knowledge-sharing forum or system they created is actively used by a wider group.
Level 1
Unconscious Incompetence
Discover & Awaken
What it looks like
Shows a less experienced colleague the tool and encourages them to experiment on their own.
How it’s confirmed
Describes pointing someone at a tool without further support.
Level 2
Conscious Incompetence
Explore & Understand
What it looks like
Provides basic instructions or examples when directly asked for help, but nothing beyond that.
How it’s confirmed
Helps reactively, only when asked, with minimal structure.
Level 3
Conscious Competence
Apply & Practise
What it looks like
Helps someone practise a structured method and gives feedback until they can produce reliable outcomes on their own.
How it’s confirmed
A real coaching example showing structured feedback over more than one attempt.
Level 4
Unconscious Competence
Intuit & Adapt
What it looks like
Diagnoses an individual’s specific development needs and adapts their coaching so the person becomes progressively more independent.
How it’s confirmed
Observed adjusting their coaching approach for someone’s specific gap rather than using a generic method.
Level 5
Mastery
Create & Transform
What it looks like
Designs development pathways, standards or learning environments that build AI capability across groups or organisations.
How it’s confirmed
A development pathway or coaching model they created is used to train people beyond those they coach personally.
See where you land.
Start with the 90-second Quick Checkup, or go deeper with the 25-question Self Assessment.