Introduction
Following a recommendation in the independent review conducted by Dame Wendy Hall FRS FREng and Jérôme Pesenti, and recently reiterated by the Government’s AI Sector Review, we have started work to develop a framework for explaining the processes, services, and decisions delivered by automated systems, in conjunction with the Information Commissioner’s Office.
Project aims
Understanding the operations and output of automated decision-making systems is a matter of ever-increasing importance. Algorithmic systems are increasingly being used in socially sensitive settings, from consumer finance to law and order. It is imperative to ensure that these systems are intelligible, both to the experts who develop and deploy them, and to the ordinary people affected by their decisions.
For that reason, the Institute welcomed the recommendation of a independent review by Dame Wendy Hall FRS FREng and Jérôme Pesenti, Growing the artificial intelligence industry in the UK, that The Alan Turing Institute work with the Information Commissioner’s Office (ICO) to develop a framework for explaining the processes, services, and decisions delivered by AI. This recommendation was recently reiterated in the government sector deal for AI.
Following these recommendations, we are working with the Turing’s academic community to generate a preliminary set of questions that such a framework will answer, and whose needs it should best serve.