Introducing GeoAI4EI: an open, trustworthy & ethical AI toolbox for epidemic intelligence
We are excited to introduce GeoAI4EI: an EU project that develops an open-source, trustworthy, and ethical Artificial Intelligence (AI) toolbox to strengthen European and global epidemic intelligence (EI) for pandemic preparedness and response.
Why GeoAI4EI?
The world is facing a growing risk of new and re-emerging infectious disease threats. Zoonotic spillovers and climate-driven disease dynamics make it clear that pandemic preparedness needs better tools:
- Faster — detecting signals earlier, when they are still actionable
- Richer — combining epidemiological, mobility, environmental, ecological, Earth Observation, and socio-economic data into one geospatial picture
- Trustworthy — AI that public health agencies, policymakers, and citizens can rely on
And it needs to be built the right way: ethics and inclusiveness are not features added at the end — they shape the methodology from day one.
What GeoAI4EI builds
GeoAI4EI builds on the Horizon 2020 MOOD platform and state-of-the-art geospatial and epidemiological modelling. On that solid foundation, the project adds:
- Multi-source data integration — six complementary data domains fused into a coherent spatio-temporal picture of disease risk
- Advanced AI — including Large Language Models (LLMs) and retrieval-augmented conversational assistants that ground answers in the governed data corpus
- Scenario-based management — tools to explore "what-if" situations and support preparedness planning
- Open, lightweight platforms — open-source and designed to run in resource-limited contexts
The toolbox is co-produced with researchers, public health agencies, policymakers, and supranational bodies — including the ECDC, HERA, and WHO — so that it solves real operational problems in epidemic intelligence.
Ethics by design
Trust in AI-driven public health tools cannot be assumed; it must be engineered. GeoAI4EI therefore commits to:
- GDPR compliance, Data Sharing Agreements, and strict anonymisation standards for all data
- Conformance with the EU AI Act, FAIR and CARE principles, and independent ethical audits
- Fairness-by-design pipelines: bias audits and sex/gender covariate analyses
- Transparency: model cards, uncertainty visualisation, and traceable sources for every answer
- Safeguards against over-reliance and misuse — all tools are decision support, never decision replacement
Inclusion & security
GeoAI4EI aims to reduce inequalities in pandemic preparedness:
- Open-source tools that any health system can adopt
- Lightweight platforms for resource-limited contexts
- Training programmes that transfer skills to practitioners everywhere
- Cybersecurity through EU-based secure hosting, encryption, and role-based access control
- Interoperability with EU infrastructures such as the EHDS and ATHINA