44% of your employees use AI — against your policies.
That is what a recent KPMG study shows. And most IT departments know nothing about it.
The phenomenon is called shadow AI: employees use ChatGPT, Claude or other tools through private accounts — without approval, governance or oversight.
There is no malicious intent.
AI is more readily available than ever: ChatGPT in the browser, Claude on your phone, Copilot in Edge. Three clicks — done.
The problem: many companies simply have not established clear ground rules. Which tools are approved? Which are not? When there is no answer, teams find their own ways.
The risk?
Sensitive data ends up in public models. Compliance breaches involving GDPR and the EU AI Act become a risk. According to IBM, one in five data breaches results from uncontrolled AI use.
What I often see with clients: IT has approved three tools. In practice, teams use twelve.
The solution is not prohibition.
The solution: build governance and define clear ground rules. Which tools are allowed? Which use cases? Who is responsible?
Where does your company stand, and which areas need priority attention? That is what my AI maturity test is for — link in the comments.
Artikel teilen↗
What does this mean for your organisation?
Let’s explore your specific challenge together.
Let’s talk↗
