Trustworthy Artificial Intelligence

Trust or trustworthiness are hard to define. There are many aspects that can increase or decrease the trust in an Artificial Intelligence systems. This is why entities such as the High-level expert group on AI (HLEG) and the European commission’s artificial intelligence act are putting forward guidelines and regulations demand trustworthiness and help to better define it. One aspect that can increase the trust in a system is to make the system more transparent. For AI systems this can be achieved through Explainable AI or XAI which has the goal to explain learning systems. This article will list some requirements from the HLEG and the European artificial intelligence act and will go further into transparency and how it can be achieved through explanations. At the end we will cover personalized explanations, how they could be achieved and how they could benefit users.