The People Problem behind the AI Paradox

Why culture, not technology, is the real bottleneck to AI adoption — and what leaders can do about it.

A few months ago, I built an entire operations system for my consultancy. Proposals, contracts, client onboarding, project tracking, invoicing. The whole thing. I’m not a developer. I didn’t hire one either. I used AI to go from idea to working system in a matter of days.

The barrier to creating has fundamentally shifted. If you can think it, you can make it. And that changes everything about how organisations can operate, especially in sectors where speed, trust, and precision matter most.

I’ve seen this kind of moment before. Working in the innovation team at one of the UK’s largest financial services firms, I watched fintech start to disrupt traditional banking products and services that hadn’t changed in decades. Small, fast-moving teams were rethinking customer experiences from the ground up. The organisations that moved quickly, and brought their people with them, created real distance from everyone else.

Now, every sector is living through that same kind of disruption. And if we’re honest about where we are, the picture is uncomfortable.

The paradox nobody can ignore

Organisations are investing more in AI than ever. Deloitte’s 2026 State of AI in the Enterprise report found that 86% of businesses are increasing their AI budgets. But only 5% are seeing substantial return on investment.

Gallup’s 2026 State of the Global Workplace report confirms this at a global scale. 65% of employees in AI-implemented organisations say it’s boosted their personal productivity. But only 12% say AI has changed how their organisation actually works. And 95% of organisations have seen zero measurable impact on profits.

Individual gains are real. Organisational gains are nowhere. That gap is not a technology problem. It’s a leadership and culture problem.

The leadership equation

Here’s what makes this moment urgent. The people responsible for leading teams through the biggest shift in a generation are themselves struggling.

Global employee engagement has dropped to 20%, the lowest since the pandemic. In Europe, it’s just 12%. But it’s the leadership picture that should be on every executive team’s agenda. Leader engagement has collapsed from 31% to 22% in three years. Leaders are reporting more stress, more anger, more loneliness, and more sadness than the people they lead.

Perhaps this is because we have leaders who acquired teams that have grown beyond their original spans of control, who never intended to be people leaders in the first place. Or perhaps leading in the middle of an organisation has become an exhausting place to be. Pressure to adopt new tools at pace, personal knowledge gaps, maybe even a little scepticism, and a weariness that gets mistakenly labelled as change fatigue.

But the data also shows us the way out. Employees whose leader actively supports their team’s use of AI are 8.7 times more likely to say it has transformed how work gets done. 8.7 times, from one behaviour change. Yet less than a third of leaders are doing that right now.

As Gallup’s CEO puts it: “The most sophisticated neural network cannot overcome an indifferent team leader.”

In a world of constant change, there is no place for change fatigue. Leaders need to build change fluency.

The human cost of getting it wrong

What happens when you deploy AI without doing the people work first? The research is starting to show us.

An eight-month study published in Harvard Business Review tracked 200 employees and found that AI users worked faster, took on broader scope, and worked more hours, all unsolicited. 83% reported their workload had increased. The researchers’ recommendation was not more technology, but what they called an “AI practice”: organisational norms around intentional pauses, sequencing, and ways of working.

Elsewhere, SHRM’s 2026 report exposed a governance gap. 92% of Chief HR Officers expect deeper AI integration, but only 25% feel their policies are future-proof. And 83% of HR leaders recognise the need for entirely new skills.

The technology is moving. The people infrastructure is not.

Create the conditions before you deploy the tools

There is a brighter signal in the data. Gensler’s 2026 Global Workplace Survey identified that 30% of workers are now “AI Power Users”, and they behave differently. They spend less time working alone, 1.5 times more time learning, and report stronger team bonds. They’re using AI in a way that is connected, collaborative, and curious.

That doesn’t happen by accident. It happens when you create the conditions for it.

Before you put AI tools in front of your teams, think about the conditions you’re creating. What’s the charter around how your team will use AI? What will you use it for, and what won’t you use it for? Where are the boundaries? How do you agree, as a team, what good looks like? In regulated industries like fintech, healthtech, legaltech, and insurtech, these questions are foundational to responsible adoption.

Right now, 78% of AI users are already bringing their own tools to work. In a regulated business, that’s not just someone using ChatGPT to draft an email. That’s unaudited decisions being made with unvetted tools in regulated processes. Shadow AI in a regulated environment is a compliance risk hiding in plain sight. The fix starts with equipping your leaders to lead the conversation, not avoid it.

Co-create, don’t deploy

You can’t mandate curiosity. You can’t force someone to find their own moment of clarity about what AI means for their work. But you can co-create it with them.

Bring your leaders into the conversation early. Let teams shape how they’ll work with AI rather than having it done to them. When people are part of building the approach, they own it. When something is handed down, they tolerate it…at best!

And here’s what I think makes this particular general purpose technology genuinely exciting. The human skills that make co-creation work, asking good questions, thinking critically, exercising curiosity, collaborating across perspectives, are exactly the skills that get better outputs from AI. How you prompt. How you evaluate what comes back. How you iterate. These are fundamentally human capabilities. In a regulated context, they’re the difference between AI that creates value and AI that creates risk.

This is about putting people at the heart of the technology. Developing a new form of literacy amongst your teams, through learning experiences and practical, hands-on building. Thinking and ways of working that stick. Not training people on tools, but building the mindset, the confidence, and the culture for people to use them well.

The best time to be curious

We’ve always had barriers to creating, to building, to those moments where something clicks. AI has lowered them to almost nothing. The opportunity for your people to be curious about their work, to find efficiencies, to shape new ways of operating: it’s never been more accessible.

Right now is the best time to be curious. But only if your culture supports it.

The question for organisations isn’t “how do we get our people to use AI?” It’s “how do we create a culture where they want to?”

This is what we do at Scail

At Scail, we believe AI capability is built through people, not tools. We help regulated SaaS businesses turn ideas into practical workflows and agents, drive real adoption, and equip leaders to make confident decisions on AI use and risk.

The people and culture work is built into our model from day one, because we’ve seen what happens when it isn’t. We surface the confidence gaps. We co-create adoption approaches with your teams. We facilitate hands-on sessions where people build with AI on real problems. And we stay until it’s working.

The result is coordinated, organisation-wide capability that delivers sustained value. Not another pilot. Not another tool rollout that fades. Embedded change that sticks.

If your AI strategy doesn’t have a people strategy inside it, it’s a shopping list. Let’s have a conversation.

Sources: Deloitte, State of AI in the Enterprise, 2026 · deloitte.com Gallup, State of the Global Workplace, 2026 · gallup.com HBR, AI Doesn't Reduce Work — It Intensifies It, Feb 2026 · hbr.org SHRM, The State of AI in HR, 2026 · shrm.org Gensler, Global Workplace Survey, 2026 · gensler.com