Artificial Intelligence in Assessing Recidivism Risk Examined Under Criminal Procedure Law and the Child's Best Interest Principle, Focusing on Legal, Ethical, and Child-Centered Implications (PID216)

1.30pm – 2pm GMT+03:00, 29 October 2025 ‐ 30 mins

Thematic Workshop Sessions

This presentation addresses the growing use of artificial intelligence in evaluating the risk of recidivism among juvenile offenders, focusing on legal and ethical implications within the framework of criminal procedure law and the principle of the best interests of the child. AI-supported risk assessment tools are increasingly being used to predict future offenses, aiming to assist judicial decision-making processes. However, when applied to children, these tools must be carefully scrutinized due to the unique developmental, legal, and psychological characteristics of minors. The study highlights key concerns, such as algorithmic bias, lack of transparency, and the limited ability of current systems to consider the social context and rehabilitation potential of children. These concerns raise questions about the fairness, accuracy, and proportionality of decisions influenced by AI in juvenile justice procedures.
 
Furthermore, the session emphasizes that the best interests of the child must be the central guiding principle in all phases of AI system development and implementation. The principle not only requires that the child's rights be respected, but also that decisions be made with a focus on reintegration and individualized treatment, rather than punitive approaches. In conclusion, while AI can provide valuable insights and efficiency to justice systems, its use in cases involving children must be strictly regulated. Transparent and accountable practices, along with child-centered safeguards, are essential to ensure that AI supports - not undermines - justice and protection for young offenders.
 
Moderated by Simon Bonk, Chair, Technology Solutions Network, ICPA, Canada