D2.1 Initial modern AI deployments and ethical gaps

This deliverable includes a comprehensive literature review that covers academic studies, international reports, and analyses related to Artificial Intelligence (AI) technologies. It explores a broad range of sectors to encompass the diverse applications of AI, with the goal of identifying and evaluating successful AI implementations and current policy developments. The identification of AI applications and their economic and societal impacts is examined, highlighting advancements, efficiency improvements, and transformative changes across various sectors. Furthermore, it assesses the capabilities and limitations of current AI tools, with a focus on technical aspects such as Machine Learning (ML) algorithms and AI models. The primary aim of this deliverable is to extract key lessons from successful AI deployments, illuminating the factors that contributed to their success while providing insights to guide future AI initiatives. Concurrently, it addresses the challenges faced during these deployments, including concerns related to bias, ethics, technical obstacles and societal implications, offering a comprehensive understanding of the challenges encountered in real-world AI applications within an ethical framework. The insights gained from the current deliverable will be integrated into the EthiTech Dialogue Hub (ETD-Hub), a discussion forum that promotes conversations on AI's legal and ethical issues among AI experts, policymakers, legal professionals, and citizens, and helps shape policies and models for the ALFIE’s AutoML platform. Moreover, the deliverable includes a comprehensive review of ALFIE's trial use cases to identify existing bias and fairness issues. It identifies state-of-the-art practices currently deployed within each use case and highlights biases inherent in these approaches. This extensive research establishes a foundational understanding of the problems ALFIE aims to address within these use cases and delineates the primary areas of focus.