Your company has AI licenses.Most of your people don't yet know what to do with them.
AI Fluency is a program and a set of products that build the capability to use AI in real work, starting with your leadership team, and measure what actually changes. It is the foundation for executing an AI strategy.
From our reference engagement, a global life sciences company
- About 15,000 employees in scope
- Full executive team engaged
- Hub, Agent and Cockpit live in production
Most organizations have rolled out Copilot, Claude or Gemini.Few can say what changed.
Usage is uneven. Training stops at awareness. Leaders cannot point to work that is done differently. The bottleneck is whether people, teams and leaders know how to apply AI to their own work, responsibly and repeatedly.
What goes in
$2.67T
Forecast worldwide AI spending in 2026
What comes out
6%
of organizations attribute significant earnings impact to AI
At the same time, about 80% of employees report personal productivity gains.
- Licensed
- Trained
- Using weekly
- Work changed
- Value proven
Common beliefs about AI fluency,and what we see in practice.
Belief: “Giving everyone an AI license makes the workforce fluent.”
In practice: A license creates the opportunity. Fluency takes skills, practice on real work and reinforcement.
Belief: “AI fluency is a skill for technical specialists.”
In practice: Every role needs it. Executives role-model, managers guide, and employees use AI to change how they work.
Belief: “HR owns the training and IT owns the tools.”
In practice: Fluency is a C-suite responsibility. The CEO, CIO, CHRO and CISO decide together what AI means for the business.
Belief: “We are teaching people how to use AI tools.”
In practice: We are teaching people how to keep learning, because the tools keep changing.
The issue is not the technology.It is whether people can apply AI to the work that matters.
AI fluency is the ability to understand, use and adapt to AI in your own work.
Most organizations have an AI strategy. Few have a plan for the people who have to execute it. Strategy, value, the operating model and technology all depend on people who use AI well.
- Strategy & DesignDefines the direction and reimagines processes and workflows.
- Value RealizationBuilds the value story and the business case.
- Operating ModelHow work gets done: structures, roles, decision rights and governance.
- Tech & DataArchitecture, AI models, agents, data and guardrails.
The layer most organizations skip
Three things we changedafter delivering this program.
- 1
Leaders learn best one-to-one, on their own work.
We planned group executive sessions. Leaders got more from individual discovery conversations and upskilling built around their own agenda.
So we now start every program with individual leader pre-meets.
- 2
Role-specific learning takes real effort, and it is where the value is.
Persona analysis is consistently underestimated in scoping. It is also what makes learning relevant to a finance analyst or a plant supervisor.
So we now scope personas explicitly and build learning paths from interviews with the people who do the work.
- 3
“Hours saved” is not a business case.
Time saved is often spent on other work, and attributing outcomes to AI is hard.
So we report four separate layers of evidence and never present estimated time as cash.
A program for leaders and employees,and the products that sustain it.
WhatWhat do we drive fluency around?
People who can apply AI in their role
- Foundations for everyone
- Learning paths built for each role, with belts
- Playbooks, challenges and quizzes on real work
WhoWho champions it?
Leaders who role-model it, champions who coach it, and learners who build the habit
- Leaders: individual pre-meets, an executive program and a 30/60/90-day plan
- Champions: managers, trainers and early adopters, with a manager toolkit
- Learners: learn, practice, apply and share what works
HowHow do we embed and sustain it?
Products, guardrails and a working rhythm that keep it going
- Engagement Hub, CAiRA the Learning Agent, and the Value & Adoption Cockpit
- Governance: guardrails, oversight and clear accountability for safe use
- Process and rituals that refresh learning as AI changes
Four layers of evidence,reported separately.
Each layer answers a different question, and each one up the stack is harder to prove. We report every layer on its own, so an early signal is never presented as a proven result.
1Literacy
Whether people are building capability, from learning completion and agreed evidence of application.
Completion does not prove mastery.
2Adoption
Whether AI is used repeatedly in relevant work, from approved usage data and participation.
Usage does not prove business value.
3Estimated productivity
The capacity benefit indicated by task evidence, surveys or agreed platform estimates.
An estimated time benefit is not a cash saving.
4Validated value
What outcome changed and why, based on a baseline, attribution and sign-off by a business owner.
Correlation alone does not show causality.
Start withyour leadership team.
An ambition conversation, individual leader pre-meets, an executive program and a 90-day plan.
From there, the program grows through scaling into an embedded capability, and onward into agentic process redesign.
See how to start