AI & Compliance · Tool

EU AI Act Classifier

Three questions about one system, and an honest answer about where it sits. Most classification tools are built by vendors and are calibrated to tell you that you are high-risk. This one will tell you when you are not, because on the financial crime side that is usually the correct answer and it is the more interesting one.

Orientation, not legal advice. Classification turns on the stated purpose of a system, which means it can be argued, and a real scoping exercise involves a lawyer reading your documentation.

1. Is the system placed on the market or used in the EU, or does its output affect people in the EU?

2. What is the system's stated primary purpose?

Does the system interact directly with a natural person, or generate content they see?