This thesis explores the transformative impact of AI in criminal trials and questions whether procedural fairness can be preserved in the automation age. This work aims to contribute to the academic debate on AI and fundamental rights, stimulate discussions concerning AI regulation at the EU level, and anticipate legal issues that member states will increasingly face in the years to come.
By focusing on the use of AI systems to generate criminal evidence, this thesis provides a reconstruction of fairness in criminal trials as a normative and legal concept and reframes the rights enshrined in Article 6 of the ECHR for the automation age (Ch. II). Secondly, it conducts a doctrinal analysis of the EU legislative measures regulating AI according to three key themes: secrecy, transparency and the reliability of AI systems (Ch. III). Finally, it shows how AI challenges the core principles of a fair trial, focusing on the redefined roles of trial actors (Ch. IV), and advances proposals to counterbalance such a transformation whilst preserving fairness in criminal trials (Ch. V). In this thesis, I contend that the use of AI systems to generate evidence in criminal trials leads to a threefold transformation: 1) new actors enter the scene, bringing commercial interests into the proceeding; 2) criminal trials become increasingly mediated by experts; 3) the traditional model of evidence examination is fundamentally changed: from inside the courtroom, audi alteram partem, to outside the trial, in absentia of the parties. To address this transformation and preserve the core principles of a fair trial, I argue for the need for legal reforms to guarantee access to information and experts for the defence. In the sector of criminal justice, a multi-level legal response, encompassing EU law, human rights law and national criminal law, is necessary to face the challenges brought by AI.