The BISTEC AI Maturity Model
Every organisation is somewhere on the same curve — from using AI for the odd task, to AI running work end-to-end. The question that places you isn't technical. It's simply: who makes the first move?
Tool → Teammate → Lead. Simple to say, hard to cross — and at every stage, a human stays in control.
Grounded in MIT's State of AI in Business, 2025 and the agentic-AI maturity work of Gartner, BCG and Microsoft.
You make the first move — every time
One question runs through it
Not how clever the model is. Not which vendor you bought. Just: when work happens, who starts it — and how much does the human still decide?
“A person asks, AI answers — you use the result once.”
Off-the-shelf chat tools, one task at a time.
A subscription and curiosity.
The stages cycle automatically — or pick one on the pyramid. Notice the control bar never drops: more initiative for AI, never less control for you.
Adoption figures: McKinsey State of AI & MIT State of AI in Business, 2025.
“A person asks, AI answers — you use the result once.”
You open Claude, ChatGPT or Copilot, ask for one thing — a draft, a summary, a slide — and use what comes back. It's genuinely useful, and it's where almost everyone starts. But it's ad-hoc, one task at a time, and you usually take the output as-is.
Off-the-shelf chat tools, used person-by-person for single tasks.
A subscription and curiosity. No engineering.
“A person presses go — AI carries the whole task through.”
You've built custom agents and skills and wired them into a real workflow. A person still kicks things off, but AI does the legwork end-to-end — pulling data, drafting, checking — and hands it back for review. This is where the value gate is: MIT found ~95% of AI efforts stall right here, trying to cross from a handy tool to a trusted, governed teammate.
Custom agents in your tools, connected to your data and people.
Engineering, integration, monitoring and governance.
“AI makes the first move and drives the work forward — people set the direction and stay in control.”
Now AI notices the trigger — a new invoice, a ticket, a market signal — and starts the work itself, end-to-end. The human's job shifts to setting the guardrails, approving the big calls, and supervising. This isn't AI let loose; it's AI-led, human-governed. Getting here takes more than tools: clear policies, trained people, and processes redesigned around the way AI works.
Autonomous workflows AI triggers, running inside human-set guardrails.
Policies, training, process redesign and strong governance.
Progress up this model isn't about handing over the keys — it's about AI taking on more of the work while people keep the final say. Even at “AI in the Lead,” humans set the rules, approve the decisions that matter, and can step in at any time. More initiative for AI, never less control for you.
How we read your stage
Maturity isn't one number. Most organisations are further along on some of these than others — and that mix is exactly what tells you where to invest next.
How much AI is part of how people actually work, day to day.
Whether AI's output is checked, and how much people rely on it.
Whether you've built your own agents and reuse them across teams.
How deeply AI is wired into your systems, data and processes.
The policies, training and guardrails that keep AI safe and yours.
Why the model matters
Off-the-shelf AI tools hit 83% adoption for quick tasks — but MIT found ~95% of AI efforts never cross from “Tool” to a trusted, integrated “Teammate,” and deliver little measurable return. The organisations that do cross it out-grow their peers. Knowing your stage is how you stop guessing and start crossing.
Source: MIT, State of AI in Business, 2025.
Free AI Maturity Assessment
Nine plain-language questions, about two minutes. You'll get your stage, a read on each of the five areas above, and the one move that matters most next. No jargon, no sales call required.
Start the assessment →Free · just your email · your result appears instantly.
Come for the free assessment. Stay for the agents. Close with governance.