A Right to Explanation for Algorithmic Credit Decisions in the UK

Abstract

Algorithmic processing of credit data is widely used in UK banks. ‘Algorithmic processing’ refers to the processing of personal and non-personal data by automated systems. This includes artificial intelligence (AI) systems such as machine learning models. The widely used description: ‘black box’ phenomenon, refers to the challenges of how and why algorithms arrive at a decision, given specific data input. The phrase ‘black box’ is used because such machine learning algorithms are very complex and constantly adapt to new input through ‘deep learning’. ‘Deep learning’ is a sub-set of machine learning. It allows computers to learn and think independently. In ‘deep learning’ technology, the decision-making process is often opaque and difficult to explain in a way that can be easily understood by humans. This inscrutability can lead to distrust in algorithms. The stakes are especially high when such algorithms are applied to decisions on access to finance because some of the most inscrutable machine learning models are used.