UNCOVER focuses on steganalysis frameworks for uncovering hidden data, while GRACE applies computer vision and NLP to detect illegal digital content.
NATIONAAL INSTITUUT VOOR CRIMINALISTIEK EN CRIMINOLOGIE - INSTITUT NATIONAL DE CRIMINALISTIQUE ET DE CRIMINOLOGIE
Belgium's federal forensic science institute, applying AI and digital forensics to law enforcement challenges including CSEM detection and steganalysis.
Their core work
Belgium's National Institute of Criminalistics and Criminology (NICC-INCC) is the federal forensic science laboratory serving Belgian law enforcement and judicial authorities. They apply advanced digital forensics, AI-based detection, and data analytics to criminal investigations — particularly in combating online child exploitation and uncovering hidden digital evidence. Their H2020 work focuses on developing tools that help law enforcement agencies detect illegal content, break steganographic concealment, and deploy AI responsibly in security operations.
What they specialise in
All three projects (GRACE, UNCOVER, STARLIGHT) involve AI-driven tools for security applications, from federated learning to adversarial AI.
GRACE specifically targets global response against child exploitation using NLP, computer vision, and federated learning.
UNCOVER addresses chain of custody for digital evidence, while STARLIGHT covers cybersecurity and emerging threats for law enforcement.
How they've shifted over time
NICC-INCC entered H2020 in 2020 with a focus on AI-powered content detection — specifically using NLP, computer vision, and federated learning to combat child exploitation (GRACE). By 2021, their focus broadened to include steganography/steganalysis and wider law enforcement AI autonomy, reflecting a shift toward deeper digital forensics and responsible AI deployment for security agencies. The trajectory shows movement from specific crime-type detection toward building comprehensive, ethics-aware AI toolkits for European law enforcement.
NICC-INCC is moving toward building sovereign, ethics-by-design AI capabilities for European law enforcement, positioning them as a key partner for future security projects that require both forensic depth and responsible AI governance.
How they like to work
NICC-INCC participates exclusively as a specialist partner, never as coordinator — consistent with their role as a national forensic laboratory contributing domain expertise rather than managing large consortia. They work in sizable networks (75 unique partners across 20 countries), suggesting they are a trusted contributor that multiple consortium leaders seek out. Their presence in both RIA and IA projects indicates they bridge research and deployment.
Despite only 3 projects, NICC-INCC has connected with 75 unique partners across 20 countries, reflecting the large-scale nature of EU security consortia. Their network spans most of Europe, indicating broad trust within the law enforcement research community.
What sets them apart
NICC-INCC brings something rare to EU security projects: they are an actual working forensic laboratory, not an academic group studying crime from the outside. This means their input reflects real operational needs — evidence handling, chain of custody requirements, and what tools investigators actually need in the field. For consortium builders, they provide genuine end-user validation that reviewers and the Commission value highly.
Highlights from their portfolio
- GRACELargest funding share (EUR 326,500) and addresses one of the most sensitive areas in digital forensics — AI-based detection of child exploitation material using federated learning.
- UNCOVERHighly specialized steganalysis project that tackles the niche but critical problem of hidden data in digital media — directly relevant to forensic evidence recovery.
- STARLIGHTLarge-scale LEA-focused AI project addressing technological sovereignty and resilience, though NICC-INCC's small funding share (EUR 32,875) suggests a focused advisory or validation role.