Orwellian odyssey: smart borders and the imperative for explainability

Sunlight, they say is the best disinfectant. Therefore, it does not come as a surprise that transparency vis-a-vis explainability of Machine Learning (ML) algorithms used in the context of border control constitutes a key societal desideratum. Integrating ML algorithms in surveillance technologies, particularly algorithms deployed in border and asylum control and management (smart borders), has transformed security measures. This has evoked profound ethical and legal concerns as a result of ML opacity. This paper seeks to contribute to the ongoing discourse on safeguarding fundamental rights in an increasingly algorithmic world, specifically in the context of migration. The paper argues that embedding explainability into AI systems used in migration control is essential to prevent the emergence of an Orwellian surveillance state, ensuring that security measures do not erode civil liberties. It proposes several technical solutions to enhance AI system explainability and address power asymmetries between affected individuals and system users, including developers. It argues that by mandating explainability of ML algorithms, we can better balance power dynamics and protect the rights of individuals impacted by AI systems.