Knowledge aWareness for emerging capabilities and risks in GPAI
Understand. Anticipate. Govern. The KnoWare Way.
About the project
KnoWare develops a prefiguration methodology and an open-access suite of tools to detect emerging capabilities and systemic risks in General-Purpose AI (GPAI) systems before they fully manifest. Grounded in Vygotsky’s Zone of Proximal Development and Explainable AI (XAI), the project moves beyond static benchmarks to provide measurable precursors, capability roadmaps, and risk-mapping frameworks that directly support compliance with the EU AI Act and the GPAI Code of Practice.
Four real-world pilot use cases validate the approach across high-impact sectors: energy and commodity trading (UC1), pandemic early warning (UC2), maritime cybersecurity and critical infrastructure (UC3), and food fraud detection (UC4). Outcomes include eight (8) Key Exploitable Results (KERs) covering a capabilities roadmap, a prefiguration methodological framework, an open-source toolbox, a GPAI benchmark suite, impact assessment methodologies, and regulatory guidance, positioning KnoWare as a direct resource for the EU AI Office, regulators, GPAI providers, and auditors.
Status:
Active from 01/05/2026
Funding Instrument:
HORIZON EUROPE
Call:
HORIZON-CL4-2025-04-DIGITAL-EMERGING-04
EC Funding:
€3,999,862.50
Duration:
2026–2029
Website:
