In this report, we analyse how Finnish law applies to the use of AI in evidence-gathering and to AI-based evidence in criminal proceedings, and whether AI could be used to assess criminal evidence. As there are no statutes specifically on the use of AI in policing or criminal proceedings, these issues are governed by more general statutes, which largely adhere to the principle of technology neutrality. Evidence law is based on the free theory of evidence, which means that AI-based evidence is generally admissible. Predictive policing systems are currently not used in Finland, but the admissibility of AI-produced predictions would in any case be limited, as evidence that is not relevant to the facts of the case should be rejected. Further, unlawfully obtained AI-based evidence should be excluded if it might endanger the right to a fair trial. AI systems are not used to assess evidence in courts, and the use of fully automated systems is not permissible.
Riekkinen et al. (Sun,) studied this question.