Belief in AI is growing, real use is lagging

AI is in the news every day. But behind the buzz sits a sober reality, and it's remarkably consistent. Two independent studies, a year apart, show exactly the same picture: belief in AI is high, structural use is lagging.

Recent research by Teamleader and iVOX (May 2026, 297 Flemish SMEs) shows that 52% see AI as a real opportunity, but only 1 in 7 (14%) use AI structurally or in an integrated way today. Nearly 3 in 4 (72%) still work mainly with manual processes and Excel. Plenty of buzz, little business so far. We picked up these insights at Work Smarter 2026 in Ghent, where this research was presented.

Last year's map still holds

That fits neatly with the five stages of AI adoption we described earlier, based on the 2025 AI Barometer of the Flemish government (2,915 companies). It showed that roughly 70% of companies are still in the early stages (no plan, plan without action, or experimenting), and only about 30% have AI really in their processes or broadly rolled out.

A year on, the picture hasn't fundamentally changed. Interest is growing, more people work with ChatGPT or Claude, but most haven't made the jump from "fun to play with" to "runs in our business every day".

Why it stalls

It stays with the individual. 51% say AI already has a big impact on the personal productivity of employees. But 55% say AI has no real impact yet on the productivity of the organisation as a whole. One employee working smartly with AI is something other than a company working smarter.

There's no framework. 6 in 10 SMEs (60%) say everyone in the company just does what they want with AI. Only 36% have a worked-out policy. Without shared instructions, integrations and agreements, every gain stays accidental and person-dependent.

The real brake isn't money, and isn't fear of jobs. Perhaps the most surprising figure: the biggest barrier is a knowledge gap (35% cite a lack of internal expertise), not cost (16%) and certainly not fear of job loss (a mere 4%). The media image of AI as a job killer doesn't match what people experience at work. And 54% simply don't know where to start.

Last year we saw the same: a large majority of companies hit the same wall, they don't have enough internal knowledge to determine their own next step, and many can't find a partner to guide them. What's missing isn't the will. What's missing is someone to take the organisation in tow.

What does work

Boring AI instead of flashy demos. The AI that really pays off isn't the impressive demo, but the reliable, everyday AI that sits in your processes and runs every day. Boring enough to trust, built-in enough to lift your margin.

From manual work to processes. Asked which processes AI should speed up, SMEs mostly name following up emails (46%), following up invoices (38%), creating quotes (35%), finding new customers (34%) and logging meeting notes (30%). All recognisable, and all things you don't want stuck with one person. We worked a few of these out in our use cases, from AI for quotes to customer service across all channels.

No jobs gone, time added. 52% already experience concrete time savings, and for those who use AI structurally it rises to 10 hours per week per employee. The question isn't whether AI replaces your people, but what you do with the time you gain. The best answer is added value: the work where your team really makes the difference. That's how you grow without constantly hiring, a story we wrote earlier in Growing without hiring.

And the barrier is lower than you think. For every AI decision, four questions come back: can I do it myself, what does it cost, what does it bring, and is it safe? A good start answers all four, small and concrete, on your own work.

The blind spots are your biggest opportunity

And here's the real opportunity. At most SMEs, AI stays stuck at writing help, translations and brainstorming. Useful, but exactly the tasks your competitor does just as well with the same tools. That's not where you make the difference.

The domains where you do stand out remain largely on the sidelines: customer service, lead generation, invoicing, and decisions based on your own figures. So an edge doesn't come from typing faster than your neighbour, but from putting AI where it really counts, wired into your processes and your own data, on the work your competitor hasn't automated yet.

How you do move forward

The common thread through both studies is clear: the belief is there, the will is there, but the knowledge and the structure are missing. That's where we come in, at your pace:

Curious what that looks like in practice? See the AI kickstart or read our approach.

The difference between a company that believes in AI and a company that gains an edge from it isn't in the tools. It's in who actually takes the step from the individual to the organisation.

Sources: Teamleader/iVOX survey, "Amai, AI?!", May 2026 (297 Flemish SMEs), and the 2025 AI Barometer, Department WEWIS, Flemish government (2,915 Flemish companies). Percentages come from those reports; the recommendations are our own.