Explainable AI

With the rapid development of Artificial Intelligence (AI) in many areas, demand for explainability and reliability of systems has also increased. AI Explainability (XAI) addresses the study of AI that can be understood by humans, since interpretability is essential for user trust, ethical, and legal issues. This article explores the intricate relations between explainability, transparency, and trust of AI systems and investigates how, on the one hand, transparency can generate trust, support ethical decision making, and engage users and, from the other, inform them about how AI systems work, in the era of XAI, drawing on extant literature and practice in the area of XAI. For this reason, the article considers the relevance of reliable AI for efficient human interaction with AI and the emergence of evidence-based perspectives for users and relevant stakeholders, respectively.