REINO UNIDO: Tratamientos de datos personales mediante Inteligencia ArtificialExplicando Machine Learning a través de la doctrina y práctica del Information Commissioner’s Office

Abstract

This article examines some of the challenges that Machine Learning techniques (ML) pose to its smooth operation in the UK from a legal perspective. Perhaps, even inadvertently to data subjects (for example, when invisible data processing occurs), ML has been widely used by the industry and the Administration. Indeed, this expansion continues with ever increasing momentum. All this notwithstanding, ML is a complex technology that is still difficult to grasp for data protection practitioners. In practice, the use of ML is frequently overshadowed by compliance concerns over, amongst other questions, the principle of transparency (strongly aligned with the explainability of ML algorithm decision-making to data subjects), privacy impact assessments, the role of the controller and the processor assessment. To cope with these issues, the UK has developed a significant «soft law» approach under the form of guidelines and other resources published by the ICO.