AI is a universally relevant research topic as well as a technology permeating our daily lives.
Smart devices monitor our doings - for safety and for other reasons. A modern car, furnished with
AI-devices, which oversees drivers’ actions, might know more about the drivers’ habits and activities
than their family, friends and neighbors.1 If read out, we could learn many things from these
devices and the IT-tools used to read them out are scientific-based and ubiquitously employed.
Therefore, AI-generated evidence should be easily transferable across borders and encounter minimal
hurdles regarding admissibility in different jurisdictions. This report tests this hypothesis by
analyzing the cross-border admissibility of AI-generated evidence. The analysis begins by explaining
the terminology and emphasizing the need for cross-border cooperation to facilitate the exchange
of evidence. Subsequently, it explores the increasing role of AI in the evidentiary process,
particularly within forensic contexts. The emergence of AI-driven evidence presents both transformative
potential and possible pitfalls, as we have observed with other science-based evidence in
the past. The report introduces real-world examples of AI-based evidence, such as DNA sample
testing, consumer product alerts, and facial recognition systems. It then delves into the specific
challenges related to the cross-border admissibility of AI evidence. Despite the absence of specific
regulations, the report finds that domestic jurisdictions possess tools to address the two main
problems: reliability of evidence proffered in criminal trials and fair trial safeguards. In conclusion,
the report advocates for the desirability of a 'universal code' to govern the admissibility of
AI evidence.