Nicole El Haber

The accelerated development of Artificial Intelligence (AI) capabilities and systems is driving a paradigm shift in productivity, innovation and growth. Despite this generational opportunity, AI is fraught with significant challenges and risks. To address these challenges, responsible AI has emerged as a modus operandi that ensures protections while not stifling innovations. Responsible AI minimizes risks to people, society, and the environment. However, responsible AI principles and practice are impacted by ‘principle proliferation’ as they are diverse and distributed across the applications, stakeholders, risks, and downstream impact of AI systems. This article presents a systematic review of responsible AI principles and practice with the objectives of discovering the current state, the foundations and the need for responsible AI, followed by the principles of responsible AI, and translation of these principles into the responsible practice of AI. Starting with 22,711 relevant peer-reviewed articles from comprehensive bibliographic databases, the review filters through to 9700 at de-duplication, 5205 at abstract screening, 1230 at semantic screening and 553 at final full-text screening. The analysis of this final corpus is presented as six findings that contribute towards the increased understanding and informed implementation of responsible AI.

Despite the increasing prevalence of artificial intelligence (AI) ethics frameworks, the practical application of these frameworks in industrial settings remains limited. This limitation is further augmented in energy systems by the complexity of systems composition and systems operation for energy generation, distribution, and supply. The primary reason for this limitation is the gap between the conceptual notion of ethics principles and the technical performance of AI applications in energy systems. For instance, trust is featured prominently in ethics frameworks but pertains to limited relevance for the robust operation of a smart grid. In this paper, we propose a lifecycle approach for AI ethics that aims to address this gap. The proposed approach consists of four phases: design, development, operation, and evaluation. All four phases are supported by a central AI ethics repository that gathers and integrates the primary and secondary dimensions of ethical practice, including reliability, safety, and trustworthiness, from design through to evaluation. This lifecycle approach is closely aligned with the operational lifecycle of energy systems, from design and production through to use, maintenance, repair, and overhaul, followed by shutdown, recycling, and replacement. Across these lifecycle stages, an energy system engages with numerous human stakeholders, directly with designers, engineers, users, trainers, operators, and maintenance technicians, as well as indirectly with managers, owners, policymakers, and community groups. This lifecycle approach is empirically evaluated in the complex energy system of a multi-campus tertiary education institution where the alignment between ethics and technical performance, as well as the human-centric application of AI, are demonstrated.