Automated decision-making (ADM) systems aim to enhance decision accuracy and fairness in areas like hiring and healthcare. Europe's GDPR and AI Act seek to ensure fairness in ADM, emphasizing transparency and human oversight, but recent failures highlight risks of bias and lack of accountability. The “right to explanation” is essential for understanding automated decisions, yet complex models often complicate clear explanations. While explainable AI methods exist, they struggle with accuracy. There's a need for ADM protections to limit fully automated decisions, especially in contexts involving human agency, to prevent unjust outcomes and maintain fairness in democratic societies.
