In GRACE (Global Response Against Child Exploitation), Web-IQ contributed NLP, computer vision, and federated learning capabilities specifically for detecting CSEM and NMEC across online platforms.
WEB-IQ BV
Dutch AI SME specializing in NLP, computer vision, and federated learning for law enforcement detection of illegal online content and cyberthreats.
Their core work
Web-IQ is a Dutch SME that builds AI-powered web intelligence tools — systems that analyze large volumes of online content using natural language processing, computer vision, and data analytics. Their real-world application focus is law enforcement: helping police and security agencies detect, classify, and act on illegal online content, including child sexual exploitation material (CSEM) and emerging cyberthreats. They bring federated learning capabilities that allow sensitive data to be analyzed without centralizing it, which is critical in cross-border law enforcement contexts. Their work sits at the technically demanding intersection of applied AI and the legal/ethical constraints of public security use cases.
What they specialise in
In STARLIGHT (2021–2026), Web-IQ is building AI-on-demand capabilities for European law enforcement to counter high-priority threats while maintaining technological sovereignty.
Federated learning was a named keyword in GRACE, indicating Web-IQ has hands-on experience enabling AI model training across distributed, sensitive datasets without centralizing raw data.
STARLIGHT keywords include ethics-privacy-and-security-by-design, sovereignty, and adversarial AI — suggesting Web-IQ is expanding into designing AI that meets European regulatory and autonomy requirements.
Data analytics appears as a foundational capability across both projects, supporting both the content detection pipeline in GRACE and the threat analysis layer in STARLIGHT.
How they've shifted over time
Web-IQ entered H2020 with a tightly scoped technical mission: applying NLP, computer vision, and federated learning to detect child sexual exploitation material — concrete, tool-level AI work in GRACE (2020). By STARLIGHT (2021), their keyword profile shifted toward systemic themes: technological sovereignty, adversarial AI robustness, human-centric design, and AI-on-demand platforms — language that signals a move from building a specific detector to building sovereign, reusable AI infrastructure for European law enforcement. The direction is clear: from specialist content analysis tool to contributor in the broader European law enforcement AI ecosystem.
Web-IQ is moving from narrow content detection tools toward broader AI infrastructure for law enforcement, with increasing emphasis on European technological autonomy and adversarial robustness — positioning aligned with the EU's push for sovereign AI in security contexts.
How they like to work
Web-IQ operates exclusively as a participant, never as project coordinator, which reflects their role as a specialist technology contributor rather than a project integrator. Both of their projects sit within very large EU security consortia — their 61 unique partners across just 2 projects indicates they are embedded in the extensive multi-stakeholder research networks typical of Horizon security RIAs. This pattern suggests they are brought in for specific AI and web intelligence capabilities, not for project management or consortium leadership.
Across just 2 projects, Web-IQ has worked with 61 unique partners in 20 countries — a notably broad network for an SME of this size, reflecting the large consortium structures common in EU security research. Their collaborations are pan-European, likely spanning law enforcement agencies, academic AI groups, and cybersecurity firms across the EU.
What sets them apart
Web-IQ occupies a rare niche: an SME with hands-on AI experience in the most legally sensitive category of online content detection (CSEM), combined with active involvement in shaping the next generation of sovereign, ethical AI tools for European law enforcement. Few small companies can claim direct project experience in both CSEM detection pipelines and adversarial AI for LEAs. For consortium builders in the EU security research space, they bring both technical credibility and an established track record of operating within the strict legal and ethical frameworks that this domain demands.
Highlights from their portfolio
- GRACEAddresses one of the most technically and ethically demanding problems in online safety — automated detection of child sexual exploitation material using federated learning — making it both high-impact and rare in the H2020 portfolio.
- STARLIGHTA large-scale, long-duration project (2021–2026) building AI autonomy and resilience for European law enforcement, with Web-IQ contributing to an infrastructure that could shape how EU police forces use AI for years.