This open-access volume examines the evidentiary use of AI in criminal proceedings, questioning whether existing criminal procedure rules are sufficient to address its unique challenges.

The widespread use of devices that collect, store, and process data about citizens' behaviour has created new opportunities for criminal justice systems to detect, investigate, and prosecute crime. Law enforcement authorities are no longer limited to searching for information or clues related to reported offences; they increasingly rely on digital investigative techniques to identify individuals, patterns, or data points that may indicate an offence. The shift from retrospective investigations toward more predictive and preventive approaches has been amplified by AI technologies using machine learning and deep neural networks.

Within AI's wide presence in the criminal justice system, this volume investigates only the evidentiary use of AI: when law enforcement authorities rely on AI for investigating and prosecuting offences and offer AI's output as evidence to tie the accused to a crime. The focus is on six countries – Germany, France, Luxembourg, the Netherlands, the UK, and the US – offering a comparative look at how each legal system handles AI evidence in criminal proceedings. Can evidence generated or processed by AI be trusted? What protections exist for the accused? The volume provides practical suggestions for law enforcement authorities and lawyers, including safeguards and limits to the use of AI in criminal cases. With the advent of regional and global efforts to regulate AI, this volume also offers solutions for supranational and national policy choices.

The ebook editions of this book are available open access under a CC BY-NC-ND 4.0 licence on bloomsburycollections.com. Open access was funded by the University of Luxembourg.