In July 2022, Eurojust and eu-LISA published a joint report and factsheet looking at how Artificial Intelligence (AI) can support cross-border cooperation in criminal justice. Issues covered by the report entail the policy and legal context for the use of AI in cross-border judicial cooperation, including ethical and fundamental rights considerations as well as relevant technologies and use-cases for the application of AI in the judicial field. For the latter, two categories of technologies are specifically explored, i.e. natural language processing (NLP) and computer vision. These two technologies are particularly relevant in applications where the processing of large-scale unstructured data is necessary. In this context, the report discusses the use of NLP applications like the following:

• Automated document processing;
• Automated translation;
• Automated summarization systems;
• NLP for evidence analysis and anonymization;
• NLP for legal research and analysis.

Furthermore, the report looks at the use of AI for forensic analysis and anonymisation of audio-visual media as well as the use of AI for the purpose of anonymisation of multi-media evidence. It pays special attention to ethical and fundamental rights considerations. The key requirements for AI systems are:

• Human agency and oversight;
• Technical robustness and safety;
• Privacy and data governance;
• Transparency;
• Diversity;
• Non-discrimination and fairness;
• Societal and environmental well-being;
• Accountability.

Although the report sees great potential for AI systems in improving the efficiency and effectiveness of the operation of judicial authorities and in reducing their costs, it also emphasizes the need for a balanced approach to ensure the protection of fundamental rights while enhancing the digital transformation process. Hence, the report recommends developing and applying a risk-based approach - integrating cost benefit analysis, fundamental rights, and data protection assessments - to determine the safeguards and deployment models for AI systems in practice.