New research from Sharp Europe suggests that, for many organisations, the cultural infrastructure supporting the open, responsible use of AI has not kept pace with the technology itself. This is slowing down the responsible implementation of AI in companies. Sharp’s research, which surveyed 2,500 SME executives in 10 European markets, shows that so-called Shadow AI, defined as the use of AI outside of approved channels or without the knowledge of colleagues, is no longer limited to junior employees experimenting with new tools. It has become a pattern at every level of the organisation, including the top.
Key findings:
- 44% of managers use AI without informing colleagues in order to gain a more competent image
- 45% of employees use AI tools without management’s knowledge
- 37% of leaders have used AI on some work without disclosing it to the team
- 33% of leaders fear they will be seen as lazy or cheating if they speak openly about their use of AI
- 31% of leaders are already concerned that untracked use of AI is a real business risk
Why this is still happening
The research points to a simple reason: organisations have invested in AI tools, but not under conditions that make their open use feel secure or supported. 35% of leaders say they lack technical assurance, 35% do not fully trust AI results and 33% fear that visible use of AI will undermine their authority.
“Leaders are silent on AI not because they are doing the wrong thing. They are silent because there is still no proper framework for action, no policies and no shared understanding of what appropriate use of AI should look like. These are not the concerns of those opposed to AI but of people who are trying to navigate this new reality without sufficient support.” – says Roland Singer, Vice President, Sharp DX Europe.
“It’s not really a question of individual behaviour. It’s about organisational culture. Companies have been quick to implement tools, but much slower to build trust and clearly define the rules that allow these tools to be used effectively.”
What comes next
According to experts at Sharp, the next phase of AI deployment will not be about access to technology. Trust, proper oversight and the creation of workplaces where the use of AI is transparent and widely accepted, rather than taking place quietly, outside of commonly agreed rules, will become key.
31% of leaders are already aware of the risks. However, the biggest problem lies in the gap between awareness and real action – this is where the real work begins. “Leaders are in the best position to give this direction,” adds Singer. “Not by having all the answers, but by talking openly about how they themselves are using AI. That changes the culture faster than any regulation.”
Sharp helps organisations securely integrate AI into everyday workflows, combining workplace technology, IT services, cloud infrastructure and cyber security to make AI implementation visible, compliant and aligned with the way people work.
The survey was conducted by Censuswide between 17 June 2025 and 25 June 2025 and involved 2,500 business owners with 50 to 250 employees in Austria, Belgium, France, Germany, the Netherlands, Poland, Spain, Sweden, Switzerland and the UK.
