All terms
Glossary
EU AI Act
The European regulation for artificial intelligence, classifying applications by risk and attaching obligations to that class.
Classification ranges from prohibited through high risk to minimal risk. For most business applications transparency obligations dominate: users have to know they are talking to a system.
Higher obligations apply in sensitive areas such as hiring, credit decisions or critical infrastructure.
How you notice it
- AI makes decisions about people.
- Customers ask how the system is classified.
- There are applications already in use.
Frequently asked
What is the first step?
A list of all AI applications in the house with purpose and affected persons. Most systems fall into the lowest tier and need little. Without that list nobody knows which applications are in scope at all, and that is what costs time later.