Study to support the technical, legal and financial conceptualisation of a European security data space for innovation

Investigations into serious and organised crime require the analysis of increasing amounts of data (e.g. from seized devices). Judicial authorities and investigators are able to use electronic data that did not exist before, delivering key evidence in concrete cases. In this context, the use of AI-driven innovative tools in line with fundamental rights enables law enforcement and other government agencies in the field of justice and home affairs to better counteract serious and organised crime. In order to test, train and validate algorithms in line with applicable legislation, law enforcement authorities and other government agencies require access to high-quality training data. The creation of a data environment would facilitate and improve the development and application of AI in law enforcement and increase the level of overall security in the EU. While crucial building blocks are available in the Member States and at the EU level, the uptake of AI and the availability of data is still limited. Within this context, there is great interest and a clear need for a European Security Data Space for Innovation ("EU SDSI"). The study outlines the high-level vision for an EU SDSI (including its potential governance model, services, use cases, and processes), and the role of the EU SDSI within the wider data sharing ecosystem (e.g. vis-à-vis the planned Europol Sandbox). The study recommends that the Commission and related EU agencies continue the development of a minimum viable product in the area of police. The EU SDSI should enable participating Member States to share non-sensitive data through a federated IT-architecture in compliance with applicable laws. A central governance model should facilitate the sharing of such data, ensure its quality, and provide dedicated services. The Commission should focus its next steps on the identification of further use cases as well as a more in-depth maturity assessment.