I think we’ve got something about AI slightly backwards. We keep talking about the technology. Which model’s best, which platform to buy, Copilot or ChatGPT or Claude or something built in-house, whether we need agents, where the data sits, whether the security controls are tight enough. All fair questions. But I’m increasingly convinced they aren’t the hard ones
The hard question is much simpler. Do your people actually want to change?
I’ve spent a lot of my career around technology change, long before AI turned up on every board agenda and conference panel. The thing I worked out fairly early is that you can buy the best technology on the planet and still get nowhere if people don’t trust it, don’t understand it, or can’t see a reason to bother.
That was true with cloud. True with automation. True with mobile working. True with just about every big shift I’ve watched. AI is only exposing the problem faster, because this time the fear is a lot more personal. People aren’t just being asked to use a new system. Some of them are quietly wondering whether the system is there to replace them. And too many leadership teams are pretending that conversation isn’t happening.
People Are Worried About What AI Means for Their Jobs. Of Course They Are.
There’s an odd thing going on inside businesses right now. Leadership talks about AI as productivity, efficiency, and opportunity. Employees often hear something completely different. They hear fewer people, lower costs, and “maybe not today, but at some point, we can do this with a smaller team.”
You can’t ignore that and then expect people to throw their arms around the technology. If someone suspects the thing you’re asking them to adopt might one day cost them their job, why would they enthusiastically walk you through every part of their role that could be automated?
Think about what we actually ask them to do. Document everything you do. Point out the repetitive bits. Show us where AI could save time. And then we act surprised when they hold a little back. That’s not resistance to technology, that’s just being human. I’d probably do the same.
This is why I think some AI programmes will fail before the technology ever gets a look in. Not because the model’s poor or the data’s bad. Because nobody dealt honestly with the people. If jobs are going to change, say so. If some tasks are going to disappear, say so. If the plan is to free people up for better work, then show them what that better work actually is. That means involving employees early, being clear about which roles and tasks may change, creating opportunities to reskill, and giving people a genuine role in deciding how AI is introduced into their work. Don’t put a smiling slide in front of them saying AI is here to empower everyone and expect it to settle a very real worry about their career. People aren’t daft. They know what efficiency means. The companies that handle this well will be the ones brave enough to have the awkward conversation instead of the comfortable slide.
We Keep Buying the Technology Before We Change the Behaviour
Here’s another one I see constantly. Company buys the platform. Training gets booked. Some people show up. There’s an internal launch, everyone gets access, and for a few weeks there’s plenty of buzz. Then, six months later, someone pulls the adoption numbers and wonders why they’re so flat.
The usual reaction is more training. I’m not convinced training is the answer. Quite often people simply have no reason to change how they work. If I’ve spent 10 years doing a task one way and that way works, why would I suddenly change because someone emailed me a link to a tool? Especially when I’m busy, my manager isn’t using it either, nobody’s checking whether I use it, and doing it properly means finding time that I don’t have to experiment, because I’m already drowning in the actual job.
We talk a lot about AI readiness. For me, this is what readiness really means. AI readiness isn’t simply having the right technology, data and security controls in place. It also means giving people the permission, confidence and time to change how they work.
Do people have permission to experiment, and the time to do it? Do they know what’s safe? Does their manager actually encourage it, or just forward the email? Do good ideas get noticed, and does anyone get credit for improving the way work gets done? Or have we just bought another tool and quietly hoped human behaviour would rearrange itself around it? It rarely does.
One thing I learned as a CIO is that you can’t just keep swapping out the technology. You have to change the organisation’s ability to change. That’s much harder, and it’s also where all the value is hiding.
Leadership Behaviour Matters More Than the AI Strategy
There’s an uncomfortable bit to this. If the leadership team isn’t using AI, isn’t learning about it, and won’t admit out loud what it doesn’t understand yet, everyone else clocks it straight away. You can’t stand on a stage telling people AI is the future and then carry on working exactly as you did five years ago. People watch what leaders do, not what’s on their slides.
If leadership still wants everything in the same format, runs the same meetings, measures the same things and rewards the same behaviour, the organisation will carry on exactly as before and AI will just sit on top of it. People might write their emails a bit faster. It might summarise a meeting. Someone might knock out a nicer deck. Useful, sure. Transformation, no.
The real shift starts when leaders are willing to question how the business itself works. Why does this approval need four people? Why does this report even exist? Why are we paying someone to copy numbers out of one system and into another? Why does this meeting happen every Friday whether there’s anything to say or not? That’s when AI gets interesting. Not “where can we put some AI”, but “why are we still working like this?” It’s a completely different conversation, and it means the leadership team has to be willing to change too, not just everyone below them.
Culture Sounds Soft Until It Costs You Millions
I know a word like culture can sound vague, and technology people often hear “culture” and picture posters on the wall and an annual survey. That’s a mistake. Organisational culture is what people actually do when nobody’s telling them what to do, and that matters enormously with AI.
If your culture punishes mistakes, nobody experiments. If it rewards people for hoarding what they know, they won’t tell you which bits of their job could be automated. If managers measure effort instead of results, people will happily hide the fact that AI got something done in an hour. If people don’t trust leadership, they’ll use AI quietly rather than openly. If nobody’s clear on which tools are allowed, shadow AI fills the gap. That’s culture, and no platform on the market fixes it.
You can buy governance tools. You can put controls in. You absolutely should – but you can’t buy trust. You have to build it, and that takes a lot longer than a procurement cycle.
The Companies Winning With AI Might Simply Be Better at Change
This is where it gets interesting. The organisations that get the most out of AI probably won’t be the ones with the biggest budgets or the flashiest technology teams. They’ll be the ones that create an environment where better ways of working can actually take hold. Where someone can say “this process makes no sense” without getting slapped down. Where a junior person can suggest a better way and actually be heard. Where a leader is comfortable saying “I don’t know, show me.” Where people can experiment safely, and a failure that was properly contained gets treated as learning rather than a hanging offence. Where improving how the work gets done is part of the job, not something you’re expected to do at the weekend.
Because the AI itself is only getting easier to get hold of. Models improve, platforms improve, tools get cheaper, and whatever looks clever today will be standard issue soon enough. Technology advantage doesn’t last. The ability to change might.
You can copy a tech stack. You can buy the same licences, run the same model and hire the same consultancy. What you can’t easily copy is an organisation where people are curious, trusted, willing to learn and comfortable questioning how work gets done. That’s where the lasting advantage might be.
It’s why, at MTI, we talk about readiness, leadership, culture and governance alongside the technology. Not because the technology doesn’t matter, of course it does, but because it’s only worth anything if something actually changes as a result.
The organisations that come through this strongest will stop treating AI as something being done to their people and start making their people part of the change.
Because this was never really a technology revolution. The technology just set it off.
The revolution is what happens to us.
This is the third article in a series from MTI’s Chief AI Officer exploring what successful AI adoption really looks like, from governance and security to agents, data readiness and measuring business value.
Next in the series: We’re Moved From ‘Wow’ to ‘How Do We Control This?’
What happens when AI adoptions moves faster than AI governance?
Continue the conversation
The questions raised in this article are ones we’re discussing with organisations every day. From AI readiness and governance to practical adoption and security, MTI helps organisations create the foundations needed to turn AI ambition into real business value.
About The Author
