Actions
DATACROS III Actions
The DATACROS project is divided into key Work Packages (WPs), each focusing on a specific objective and together ensuring a comprehensive approach to the project’s goals, implementation, and impact.
1. New Data Sources and Modules
Focuses on expanding the DATACROS toolbox by integrating new financial and asset-related data sources. The aim is to provide investigators with a comprehensive and interconnected data ecosystem for enhanced risk assessment and financial investigations. New data will include:
- Beneficial ownership (BO) and business registers
- Company data
- Real estate data
- Public procurement and funds data
- Virtual currencies and transactions
- Satellite imagery
API/web-service connections will guarantee real-time data access.
2. Next-Generation AI Tools and Analytics
Develops and implements advanced AI and machine learning models to improve the detection of high-risk entities and transactions, while minimising false positives. It enhances anomaly detection, entity resolution, and natural language processing (NLP) capabilities, allowing for automated identification of organised crime (OC) infiltration patterns across financial domains. Main features include:
- Machine learning-empowered anomaly and risk detection models
- NLP-based analysis of information from unstructured data sources
- Processing of satellite images and videos
- AI-driven anomaly indicators to detect illicit financial activities
3. Piloting in Operational Use Cases
Tests the toolbox in real-world investigations to assess its effectiveness in tracing criminal assets and financial flows. End-users – including law enforcement agencies (LEAs), financial intelligence units (FIUs), anti-corruption authorities, and investigative journalists – will apply the tool in selected cross-border financial crime cases. Key outputs:
- Four pilot investigations in areas such as asset tracing, OC financial crimes, public procurement, and subsidy fraud
- Reports on case study results and tool effectiveness
- Recommendations for improving investigative workflows
4. Multi-Stakeholder Cooperation and Training
Provides targeted training and knowledge exchange for LEAs, FIUs, asset recovery offices (AROs), and investigative journalists. The goal is to enhance awareness and practical skills in using data-driven tools to combat financial crime and organised crime infiltration. Activities include:
- Four training sessions on data-driven OC investigations
- An international conference on financial crime and AI in Bucharest
- Online knowledge-sharing platform for investigators and journalists
5. AI: Ethics and Legal Compliance
Assesses the risks, benefits and regulatory implications of using AI for financial crime investigations and ensures that the development of the DATACROS tool is aligned with the EU AI Act and the GDPR. Contributors include experts from law enforcement, academia, and policymaking. Key outputs:
- Establishment of an Advisory Board
- White paper on AI compliance with EU data protection laws
- Policy recommendations for law enforcement and financial regulators
6. Communication and Dissemination
Ensures the visibility of the work done within DATACROS III by promoting its results and demonstrating its real-world practical applications. To this end, journalists from partner organisations will use the toolbox to report on OC-related financial crime cases, thereby raising public awareness of the impact and relevance of the tool. Key outputs:
7. Management
Ensures the smooth execution of the project by coordinating the work of all partners and monitoring progress against key performance indicators (KPIs). It includes:
- Financial and administrative oversight
- Quality control and risk management
- Organisation of internal meetings and reports
