AI Model Risk Management is particularly at home in the areas of artificial intelligence, big data, smart data and automation. Companies are increasingly using AI models to automate processes or support decisions. However, this gives rise to specific risks, such as errors in the model, distorted results due to poor data or uncertain decisions.
AI Model Risk Management describes all the methods and measures that companies use to recognise, assess and manage these risks. The aim is to use AI models as safely and reliably as possible. This is achieved through thorough testing, regular checks of the data and clear guidelines on how the models may be used.
A simple example: a bank uses an AI model to decide who to lend to. However, if the model only recognises incomplete or incorrect data, good customers could be rejected. With AI Model Risk Management, the bank regularly checks the model and ensures that it is working fairly and sensibly. In this way, AI Model Risk Management protects companies from making costly mistakes and losing the trust of their customers.
Deutsch 
English (UK)