AI Adoption: why businesses are stuck at Level 1

I keep coming back to a stat from PwC's 2026 Global CEO Survey. 56% of CEOs say AI has delivered neither cost savings nor revenue growth. Really... nothing? And BCG found something similar - 74% of companies struggling to achieve and scale value from AI.

I don't find those numbers surprising. I find them useful. Because I think they're telling us something specific, and it's not that AI doesn't work.

It's that most businesses adopted it at Level 1, expected it to deliver like Level 5, and gave up when it didn't.

Level 1 was never going to change your business

So here's what Level 1 actually looks like. Someone, usually in leadership, decides the company needs to be doing AI. Someone in IT buys the licences, someone in leadership sends the all-staff email. People log in, experiment a bit, and get on with their day.

The first three months are underwhelming. The gains are real but modest, kind of absorbed by the time it takes to get comfortable with something new. Leadership looks at the numbers, sees little movement, and draws what feels like the obvious conclusion: this isn't working for us.

I think that conclusion is almost always wrong. But I also understand why it gets made.

The problem is that giving an employee a Claude or ChatGPT account is the starting line, not the strategy. It's a bit like handing someone a laptop and wondering why productivity hasn't transformed. Access isn't adoption. And adoption isn't impact.

The five levels - where does your organisation sit?

I want to be clear about what I mean by "levels" here, because it's easy to read this as a comment on the technology. It's not. The technology available at Level 1 and Level 5 is largely the same. What changes is how deeply it's embedded into how your organisation actually works.

All five levels deliver value. I think that's important to say. But the nature of that value, and the scale of it, is very different depending on where you are.

Level 1 - Chat. People have access and they're using it. Asking questions, speeding up tasks, getting help with things that would have taken longer before. Real value, but contained to whoever's using it at any given moment. Nothing about how the business operates has changed.

Level 2 - Skills. Some individuals get genuinely good at it. They develop repeatable ways of working with AI - workflows and approaches that save them real time. This is still individual though. The value doesn't travel. When that person takes a week off, the efficiency goes with them.

Level 3 - Shared Skills. This is where things start to get interesting. One person's workflow starts spreading across the team. What used to take hours becomes something everyone does faster. Knowledge stops being personal and starts being organisational. This is where the multiplier effect starts.

Level 4 - Integrated Workflows. AI is built into how the business runs, not bolted on. Processes are redesigned around what AI makes possible. You're not saving time on old processes - you're running better ones.

Level 5 - Autonomous. AI acts on behalf of the business. Agents handling tasks end-to-end, without someone prompting every step. The organisation gets leverage it couldn't have had before.

Why the numbers change at Level 3

Another piece I found interesting was this CIO.com article that pulls together research from across the leading consulting firms. One line really captures how I feel about it: "Tools don't pay off until work changes." I think it's exactly right.

Levels 1 and 2 make individuals faster. That's genuinely useful, but it doesn't show up clearly in the business numbers. The gains are real but scattered, hard to point to. Easy to dismiss.

Level 3 is different. When one person's better way of working becomes the team's better way of working, the maths changes. One investment in figuring something out multiplies across the organisation. That's the moment AI starts to feel like a real business advantage rather than a personal productivity tool.

So when I look at those PwC and BCG statistics, I don't think those businesses failed at AI. I think they measured their results at Level 1, found them underwhelming - which they would be - and stopped. They were right that Level 1 was underwhelming. They were wrong to take that as the verdict.

The gap between Level 2 and Level 3 isn't a technology problem

This is the part I think gets underestimated. When businesses struggle to move up the levels, the assumption is usually that it's a technology problem. In my experience, it almost never is.

The gap between Level 2 and Level 3 is a people problem. Attitudes, habits, trust - and the change management challenge of getting a team to genuinely shift how it works together. That's a harder thing to solve than deploying a new tool, and it doesn't get easier if you just wait.

Here's something I've noticed though. In almost any organisation of 30 or 40 people, there are four or five individuals who've already quietly figured it out. They've moved themselves to Level 2 or beyond. They have workflows that save them hours. They mention AI in passing and make their colleagues curious.

I think the most underrated move is getting those people to share what they've found - not in a training session or a policy document, but in the language of the actual work. "Here's how I write a brief now. Here's how long it used to take. Here's what I do differently." When that story connects to a problem someone else in the room actually has, something shifts.

It's that combination - the technology, the changed way of thinking, and a visible link to real business value - that moves an organisation from Level 1 to Level 3. Not the tool itself. The story around it, and the people willing to tell it.

Level 1 is a beginning, not a verdict

Most businesses that have walked away from AI didn't fail at AI. They started, measured too early, and stopped. Level 1 is where you learn. It's not where the value lives.

Getting from there to Level 3 and beyond is as much about people as it is about tools. The organisations that work that out - that invest as seriously in the coaching, the storytelling and the cultural shift as they do in the technology - are the ones whose numbers start moving.

If you're at Level 1 and wondering why it isn't working yet, I'd say: you're asking the right question. Give me a shout if you want to think through where you are and what the next step looks like.