AI adoption starts with people, not with tools
Most AI projects stall on people, not on technology. Adoption needs three things, in this order: a change in mindset, real space in people’s working week, and a clear structure with rules, data and owners.
Are people ready for AI?
There is openness, but it comes with conditions. In a European Commission survey, 67.4% of Romanian respondents support integrating AI into public administration within the next 12 months. Only 52.6% believe AI will make the world a better place. The main concerns are cyber attacks, the removal of human judgement and a loss of trust.
In companies, the fears are similar. In the SME Digital Index 2025, the digitalisation risks with the highest scores are data loss and vulnerability to cyber attacks, followed by difficult integration between solutions and a lack of digital skills. Software use and artificial intelligence are tied as the areas where entrepreneurs most often say employees need training (31.26% each).
Why does AI adoption stall in a company?
- Mindset. AI is seen either as a threat to jobs or as a toy for the tech enthusiasts.
- No space. People get new tools on top of an already full week. They have no time to learn them properly.
- No structure. Everyone uses whatever they like, with no rules for data and no measure of results.
In the Brand–Culture–Performance model, all three are organisational friction and cultural alignment. A better tool does not fix them.
How do you create space and structure for AI?
- Say why. Link AI to the company’s strategy and say clearly what will not happen. Mindsets change first in the leadership message.
- Protect time. Hours each week for learning and experiments, protected from urgent work.
- Set simple rules. Which data can go into AI tools and which cannot. Who approves a new tool.
- Start small. One or two processes, an owner for each, and a measure before and after.
- Leave the decision with people. Say explicitly that AI prepares and people decide. It answers the fear about human judgement directly.
- Train on real problems. Exercises on everyday work, not general courses about AI.
How ready is your company?
- Leadership can say in one sentence why the company uses AI.
- People have time set aside for learning, not just new tools.
- There are written rules on which data can go into AI tools.
- Every AI project has an owner and a measure of its result.
- It is clear who makes the final decision when AI makes a recommendation.
From my experience
In a project with a services company, the team of specialists resisted the move from handling incoming requests to contacting customers proactively. People saw the new initiatives as yet more tasks, without understanding enough of the economic context that made them necessary. That was what the change hinged on: the managers’ ability to explain why, to clarify expectations and to support the team in the new way of working.
I consider it an essential lesson for AI adoption too: before you ask people to work differently, help them understand what problem the change solves and what it means for them in practice. Leadership’s enthusiasm does not transfer to the team automatically.
Frequently asked questions
Where do I start implementing AI in my company?
With the leadership message and one or two processes, each with a clear owner. The tool is chosen after that.
Why are employees afraid of AI?
Of losing their jobs, of making mistakes and of their judgement no longer counting. Each fear has a concrete answer: rules, time to learn and decisions left to people.
What rules does a company need for AI?
At least three: which data can be used, who approves new tools and who is accountable for the result of each project.
Sources
- Rodriguez Müller, Tangi, Lasio, Where people accept AI in government, and where they stop, European Commission, Joint Research Centre (forthcoming). Comparative survey across several member states; not statistically representative, quota sampling. Presented at the Transylvania Digital Innovation Summit, Cluj-Napoca, 8 October 2026.
- Florin Jianu, Dragoș Pufulete, Elena Dascălu, IMM Digital Index 2025: Carta Albă a IT-ului (SME Digital Index 2025: The IT White Paper), National Council of Small and Medium-Sized Private Enterprises in Romania, 2025. Presented at the Transylvania Digital Innovation Summit, Cluj-Napoca, 8 October 2026.