“I just feel like I’m not really a part of it, like I’m sitting there watching a YouTube video or something.” This is how a child in Ireland described their own hearing, held by video link.
The quote opens a new policy briefing for the European Union and the Council of Europe, researched and drafted by Cédric Foussard and Mariana Pérez Cruz from ICJA for Penal Reform International as part of the EU-funded PPRO-Child project. The briefing Children in Digital Justice: Protecting Procedural Rights from Digital Offences to Remote Hearings examines how the digitalisation of criminal justice in Europe affects children who are suspects or accused persons. It identifies two areas where the current EU legal framework fails to provide adequate protection: digitalised proceedings, and offences allegedly committed by children through digital technologies.
The analysis draws on empirical research carried out in Ireland, Romania, Spain and Ukraine by Penal Reform International and Terre des hommes. This research includes rapid needs assessments and contributions from Child Advisory Boards made up of children with lived experience of justice proceedings. It also draws on judicial perspectives gathered across eight jurisdictions, and on recent work by the EU Fundamental Rights Agency, the Lanzarote Committee and the UN Special Rapporteur on the independence of judges and lawyers.
A legal framework designed for analogue proceedings
Directive 2016/800 on procedural safeguards for children who are suspects or accused persons remains the principal binding instrument in this field. It was drafted for in-person proceedings, and it does not specify how its core guarantees apply when proceedings are conducted remotely. Those guarantees include the right to effective participation (Article 16), individual assessment (Article 7), confidential legal assistance (Article 6) and protection of privacy (Article 14). The Artificial Intelligence Act (Regulation (EU) 2024/1689) classifies risk-assessment tools used in criminal proceedings as high-risk, but it establishes no child-specific requirements. As a result, no EU instrument currently governs how children’s procedural safeguards operate when a child appears by videoconference or when an algorithm informs a decision on detention.
Remote hearings: presence without participation
The PPRO-Child findings indicate that remote hearings substantially weaken children’s capacity to participate effectively. Nearly a third of the children surveyed entered remote hearings without knowing their rights or understanding what was about to happen. In Spain and Ukraine, fewer than 35% felt able to communicate privately with their lawyer during a remote hearing. Of eleven children consulted at Oberstown Children Detention Campus in Ireland, all but one reported having no understanding of what was happening, and several described themselves as spectators rather than participants. Nine in ten legal professionals surveyed considered that remote hearings reduced their ability to establish a connection with the child.
The briefing also examines the effects of the digital divide and the particular risks faced by neurodivergent children and children with communication difficulties. These children make up a significant share of those in contact with justice systems. In consequence, the briefing argues that physical hearings must remain the norm. Remote hearings should be limited to procedural and administrative matters and excluded from any hearing that determines guilt or affects a child’s liberty. Where they are used, they should require the child’s freely given and informed consent, obtained in consultation with a lawyer, together with a prior individual assessment confirming that the format serves the child’s best interests.
Artificial intelligence and automated decision-making
Algorithmic risk-assessment tools are increasingly used to inform decisions concerning children. Because these tools are trained on historical criminal justice data, they may reproduce existing patterns of over-policing and structural disadvantage. The current framework does not determine how AI-generated outputs relate to the individual assessment required under Article 7 of Directive 2016/800. It therefore does not prevent a situation in which a child assessed as low risk by a trained professional is nonetheless detained on the basis of an algorithmic output. The briefing concludes that decisions concerning a child’s liberty, guilt or sentence must be taken by a human professional trained in child justice, on the basis of an individual assessment and with full due process.
Children suspected or accused of digital offences
European frameworks on online harm have been developed primarily around the protection of child victims and the prosecution of adult perpetrators. The briefing shows that they offer limited guidance where the person suspected of an offence is a child. It recalls that such a child remains a child in conflict with the law, entitled to the full protection of Directive 2016/800, Articles 37 and 40 of the UN Convention on the Rights of the Child, and international child justice standards. Empirical evidence shows a significant overlap between online aggression and victimisation among children, as well as marked gender differences in how image-based cases are handled. Moreover, the Lanzarote Committee’s 2026 compliance report found that almost two thirds of the States examined still permit the prosecution of children for possessing their own self-generated images. Together, these findings point to a real risk of over-criminalisation.
Recommendations
The briefing sets out seven recommendations. Five are addressed to the EU institutions:
- Amend Directive 2016/800 to establish binding standards for digitalised proceedings involving children.
- Prohibit AI-led decision-making in child justice and introduce child-specific requirements for other high-risk AI systems.
- Clarify the treatment of AI-generated content in the revision of Directive 2011/93/EU while preventing the over-criminalisation of children.
- Promote multi-agency safeguarding structures for digital peer-to-peer harm.
- Extend data collection under Article 21 of Directive 2016/800 to digitalised proceedings.
Two are addressed to the Council of Europe:
- Issue interpretive guidance on the 2010 Guidelines on child-friendly justice covering digital proceedings.
- Support consistent national implementation of the June 2026 Lanzarote and Budapest Committees’ joint statement on AI-generated content.
Full citation
Penal Reform International, Foussard, C. & Pérez Cruz, M. (2026). Children in digital justice: Protecting procedural rights from digital offences to remote hearings. A policy briefing for the European Union and the Council of Europe.
Read the full briefing:





