There's a conversation happening in most organisations right now that nobody's quite having out loud. AI tools are being used, results are being delivered, but the how is staying quietly in the background.
We talk a lot about employees adopting AI without telling their managers. It's been a recurring headline for a couple of years now. When we surveyed 2,500 SME leaders across Europe recently, what stood out wasn't what was happening at the bottom of the org chart. It was what was happening at the top.
In fact, 44% of the business leaders we spoke to said they use AI to help them sound more knowledgeable at work, without telling colleagues. 37% have used AI on a piece of work and kept quiet about it. A third worry that being open about using AI would make them look lazy, or like they're somehow cheating.
What’s surprising wasn't the scale of those numbers. It was what they say about where we actually are with AI culture in the workplace.
The tools are there. The culture is catching up
Most organisations have spent the last few years focused on AI access. Which tools to use, which teams to train, how to stay ahead of competitors. That investment has paid off in some ways, with 70% of the senior leaders in our research saying they actively use AI at work.
But access and openness aren't the same thing. And right now, there's a gap between the two.
When leaders keep their AI use quiet, it's usually not because they're doing anything wrong. It's because the culture around AI hasn't caught up with the technology. There's no shared language for what good AI use looks like. No real clarity on what's encouraged, what's acceptable, and what requires disclosure. So people default to keeping it to themselves.
That's understandable, and it's exactly what needs to change.
It's also worth noting that the concerns leaders have around AI aren't that different from those felt across the rest of the business. 35% of the leaders we surveyed say they lack the technical confidence to use AI fully. Similarly, 35% don't entirely trust AI outputs. These aren't the reservations of people who are disengaged from the technology. They're the reservations of people who are already using it and want to use it better. That distinction matters, because it changes how organisations should be responding. Less policing, more support.
The quiet use becomes the bigger risk
When AI use happens without visibility, everything that should come with it gets left out too. Oversight, governance, data security: none of that can function properly if nobody knows which tools are being used, by whom, or for what.
The risks are real and specific. Employees and leaders using unsanctioned AI tools may be sharing sensitive business information with third-party platforms, often without realising it. Data that feels like a quick query can end up stored on external servers, outside of any compliance framework the business has in place. For SMEs operating under GDPR and increasing scrutiny around data handling, that's not a theoretical risk. It's a live one.
Nearly a third of the leaders we surveyed are already worried that untracked AI use could have a negative impact on their business. And yet the behaviour continues to grow, at every level of the organisation.
The answer isn't stricter rules or blanket bans. In our experience, those tend to push the behaviour further underground. What works is making it easier and safer to use AI openly than to use it quietly. Clear policies. Secure, sanctioned platforms. Training that builds real confidence rather than just ticking a box.
The opportunity for leaders
The fact that senior leaders are feeling the same cultural friction as their teams is not just a challenge to address. It's also a starting point.
Leaders who are open about how they use AI, including the areas they're still navigating, do more to shift workplace culture than any policy rollout. It signals that AI is a tool, not a shortcut. That using it well is a skill worth developing, not something to keep to yourself. When that message comes from the top, it tends to stick.
Building that kind of environment takes clear guidance on which tools are approved and why. It takes training that goes beyond the basics and builds genuine confidence. It takes a governance framework that gives people guardrails without making AI feel like a compliance exercise. And it takes leaders who are willing to model the behaviour they want to see across their teams.
That shift doesn't require having all the answers. It just requires being willing to have the conversation.
At Sharp, that's exactly what we help organisations do. Build the foundations for AI adoption that's visible, secure and built around how people actually work. Because the goal isn't just more AI. It's better AI, used with confidence, at every level of the business.
For Sharp Europe's latest thinking on AI confidence and adoption, watch our recent discussion on agentic AI with Quocirca here.
