The Three Levels of AI Fluency.
A practical framework inspired by Reid Hoffman's three levels — fluency as observable behaviour, not adjectives.
Behind every AI Fluency Score in HRPulsar sits the same scale: five behavioural dimensions, each a ladder of four steps, summed into three levels. This page is the framework itself — where it comes from, the dimensions, and a five-question self-assessment.
Where the framework comes from.
Reid Hoffman — co-founder of LinkedIn, partner at Greylock, author of The Start-up of You and Impromptu — described three levels of AI fluency in early 2026: from voice-driven chat assistants, through persistent agents, to multi-agent orchestration. The idea that stuck is that fluency is behavioural: defined by what people do, not by how confident they feel.
“There will no longer be solo specialists working alone. Each of us will operate together with our own set of AI agents.”
The tools moved fast after that. Voice, chat-first research and connected data went mainstream; agents that run whole tasks arrived; and the EU AI Act now expects organisations that use AI to keep their people AI-literate. So we kept the three levels and rebuilt the markers underneath around what separates practitioners now: what you hand over, what the AI knows, what it can reach, how you check, and where you stop.
No vibes. Just behaviour.
Five dimensions, each a ladder of four steps. Pick the step that describes you, add the five up, and the total maps to a level. The steps are observable: a colleague could watch you work for a week and place you. The same scale is designed to feed the team-level AI Maturity Index inside HRPulsar — your individual level is built on the scale your future employer will measure.
Basic
0–5
AI as a smarter search box. Answers and drafts from what you paste in; the work, the checking and the judgement stay manual.
Middle
6–11
AI does whole tasks with standing instructions and connected data. You check the output against a rubric and know where a person must decide.
Advanced
12–15
Agents run recurring work inside your systems, within limits you set, measured over time, and explainable to an auditor.
01. What you hand over
How much of a task the AI does before you touch it.
Step 0I ask questions and get answers or ideas. The work itself is still mine.
Step 3Agents run recurring work without me starting it. I review the exceptions.
02. What the AI knows
What is already in place before you type.
Step 0I type a fresh prompt every time.
Step 3I write instructions my team reuses, and I version them like any other document.
03. What the AI can reach
What it can see and do beyond the chat window.
Step 0Only what I paste into the chat.
Step 3Agents can act in those systems — create, update, send — within permissions I have set.
04. How you check
How you know the output is right.
Step 0I trust the output if it looks right.
Step 3I track quality over time — samples, error rates — and change the setup when it drifts.
05. Where you stop
What stays out of AI tools, and where a person decides.
Step 0I am not sure what I am allowed to put into AI tools.
Step 3I can explain to a candidate or an auditor how AI was used in a decision, and it is documented.
Shown here: the two ends of each ladder. All four steps of every dimension are the assessment below.
Where are you on the scale? Rate in 2 minutes.
Self-assessment across five behavioural dimensions. Computed entirely in your browser — no account, no save. You'll get a level, a profile across the five dimensions, and up to three concrete next moves.
Apply the scale to your entire team.
In developmentThe scale is designed to run inside the platform on the same evidence the assessment modules already collect, producing an AI Fluency Score for individuals, teams, departments — and the company as a whole, broken down to the same five dimensions you just answered for yourself. Assessments, competencies and grading ship today; the fluency scale on top of them is being built with our design partners, and what we do and do not measure to get there is written down in full.
Share your level. Reserve a verified profile.
Self-assessment is the starting point. A verified, portable AI Fluency profile — backed by external review — is what makes fluency credible across employers.
Get your team's collective AI Fluency Score.
HRPulsar Cloud gives you the team view, the trend lines, and the supervision-gap report — same methodology, all employees.