View of An Architecture to Quantify the Risk of AI-Models | Proceedings of the Upper-Rhine Artificial Intelligence Symposium