AI Fundamental Rights Impact Mining

The EU’s AI Act brings a range of A- based products into the scope of EU product regulation. It places a set of risk management and monitoring requirements that expand EU health and safety and consumer protection legislation by expanding the scope to protect Fundamental Rights [1]. A major challenge is the lack of an evidence base or case law where AI applications are judged to have impacted a citizen’s fundamental rights. Although databases have been assembled that track ‘AI incidents’ as reported in news articles [2], their categorisation of AI risks based on the impacted fundamental rights is limited, e.g. see the MIT domain taxonomy [3]. This project will develop a classifier that can characterise existing AI incidents against EU fundamental rights breaches. It would aim to catalogue incidents as AI risk assessments using existing models of AI risk classification [4][5] but with the goal of providing a resource for public bodies involved in future Fundamental Rights Impact Assessments as required by the AI Act.