AI use is already happening in most businesses. Someone is drafting emails with ChatGPT. Someone else is testing Copilot. A manager may be using AI to think through reporting or strategy.
But in most teams, that knowledge stays with the individual. Useful shortcuts get trapped in someone’s head. Prompts get rewritten from scratch. Quality depends on who happens to be using the tool.
The 10 Levels of AI Mastery explain the difference between casual AI use and real business capability. They show how people move from early experimentation to practical, governed and repeatable team capability.
Capability is not just what one person can do with AI. Capability is what the business can repeat safely when that person is not in the room.
Each stage marks a genuine shift in how people work with AI and what the business can do because of it.
Literacy
Levels 1–3
People are learning how to use AI properly
They can ask better questions, write clearer prompts and produce useful first drafts
Fluency
Levels 4–6
People can use AI repeatedly in real work
They build reusable prompts, templates, assistants and workflows for common tasks
Mastery
Levels 7–9
AI starts supporting work across teams and systems
Teams connect workflows, manage handoffs, reduce manual rework and design more advanced AI-supported processes
INGRAINED
Level 10
AI capability is embedded in the operating model
Leaders govern AI use, review performance, protect judgement and make sure people grow as the technology improves
For most businesses building AI capability, Levels 1 to 6 are the practical working range.
The first meaningful gain usually comes from moving people out of casual experimentation and into Level 3 – Craftsperson, where they can use AI reliably for repeated work.
The next major shift comes at Level 6 – Orchestrator, where individual capability becomes shared team capability. That is where AI stops being something one person uses well and starts becoming part of how the business works.
Click any level to see what it looks like in practice. Explore what people can do, how their thinking shifts, what they build, and what governance the business needs as capability grows. Toggle between SMB and Enterprise to see how governance scales. The discipline is the same. The structure fits the size of the business.
A practical map of how AI capability develops inside a business, built on the INGRAIN AI methodology.
Toggle between SMB and Enterprise to see how architecture and governance scale to the size of the business. SMB governance is not weaker governance; it is the same discipline applied through simpler ownership and practical controls.
methodology, applied by Dovetail Digital.
The 10 Levels show how AI capability develops.
Your AI Impact Report shows where your team sits today. It estimates:
The report turns the framework into specific numbers and priorities for your business.
The AI Impact Report gives you a practical view of:
Use the sections below to understand why the levels matter, how they connect to business value, and how Dovetail applies the INGRAIN AI methodology.
A business can buy AI licences and still see limited return if people do not know how to use them well.
Self-serve training can give individuals useful information, but it often fails to create shared language, common processes or transferable ways of working.
The goal is not to turn everyone into an AI expert. The goal is to build enough shared capability that AI becomes a practical, governed and repeatable part of how work gets done.
Dovetail uses the INGRAIN AI methodology to help businesses turn access to AI into practical capability.
That means building:
In simple terms: Tools give people access to AI. Dovetail, using the INGRAIN AI methodology, helps a business turn that access into practical capability.
How capability develops
The INGRAIN AI methodology recognises that AI adoption follows a predictable progression. Skipping stages increases risk and resistance.
| Stage | What it solves |
|---|---|
| Hype | Confusion, fear, and over-expectation about what AI can do |
| Habit | Inconsistent or shallow AI use across the team |
| Discipline | Unreliable or unsafe behaviour as AI use becomes regular |
| Governance | Uncontrolled scale as AI starts affecting more of the business |
| Defensibility | Trust, accountability, and assurance once AI is embedded in operations |
The 10 Levels of AI Mastery sit inside this larger progression. The early levels build habits. The middle levels build discipline. The later levels add governance. The endpoint is a business that can explain, defend, and sustain its AI decisions.
The levels are not a pass or fail score. They are a practical diagnostic. They help answer:
The important shift is from individual AI use to organisational AI capability.
| Early AI use | Structured AI capability |
|---|---|
| One person experiments | The team has shared methods |
| Results depend on who uses the tool | Outputs become more consistent |
| Prompts are rewritten from scratch | Prompts and workflows become reusable |
| Knowledge stays in someone’s head | Knowledge becomes transferable |
| Governance is informal or unclear | Safe-use rules and human review are built in |
| Value is hard to measure | Time savings and capacity gains can be tracked |
Your AI Impact Report gives you a practical view of:
No. Most businesses do not need developers to begin building AI capability.
The first stage is usually about helping people use AI more effectively in everyday work: drafting, summarising, planning, analysing, documenting, communicating and improving recurring tasks.
It depends on where the team is starting from.
Some businesses can make useful progress quickly by moving from casual experimentation to shared prompts, templates and safe-use habits. That is often where the first practical gains appear.
The AI Impact Report gives you a clearer view of where your team sits today, how consistently AI is being used, and where structured uplift could create practical value.
From there, the next step depends on the result.
For some businesses, the right starting point is foundational training and simple guardrails. For others, it may be a more focused activation plan around priority workflows. More mature teams may be ready to build reusable assistants, internal processes or more structured AI-supported systems.
