The availability of big data, combined with advances in Artificial Intelligence (AI), offers extensive capabilities to public and private actors. However, cross-organisation and cross-border data sharing while respecting current GDPR laws and regulations becomes increasingly difficult, as such information involves the collection and cross-referencing of multidimensional personal data, often of very low granularity.
HARPOCRATES leverages new cryptographic schemes to enhance the capabilities of Privacy Preserving Machine Learning (PPML) and Federated Learning (FL), thus enabling decentralized training, validation and prediction on encrypted data. These privacy preserving and secure computing services allow users to benefit from cloud-based artificial intelligence while maintaining control over their data.
HARPOCRATES will enable the demonstration of digitally blind evaluation systems in practical application scenarios, helping to build fairer, more democratic and unbiased societies.
This project has received funding from the European Union’s Horizon Europe research and innovation programme under Grant Agreement No 101069535.