The University of Amsterdam (UvA)is a top-tier research institution in the Netherlands. Through its Faculty of Science and the Parallel Computing Systems (PCS) group, UvA contributes to the LICORICE (reLIable and sCalable tOols foR self-sovereIgn identity and data proteCtion framEwork) project by developing secure and scalable AI systems. Its research focuses on privacy-preserving machine learning, secure computation, and high-performance computing. UvA leads the research and development of efficient neural network inference methods using advanced cryptographic techniques such as Secure Multiparty Computation (MPC) and Homomorphic Encryption (HE). These solutions are GDPR-compliant and tailored for real-world applications. UvA’s work reinforces the project’s scientific foundation and enables practical, privacy-by-design systems in critical sectors like healthcare.
Role in LICORICE'S PROJECT
The University of Amsterdam (UvA) contributes to LICORICE by advancing the scientific and technical foundations of privacy-preserving neural network inference. UvA analyses how cryptographic tools like Secure Multiparty Computation (MPC) and Homomorphic Encryption (HE) affect the speed and privacy of secure AI systems. Through systematic evaluation, UvA identifies key performance bottlenecks and proposes novel improvements that reduce computational and communication overhead.
UvA further designs and implements optimized privacy-preserving inference solutions adapted to real-world requirements, especially in sensitive domains such as eHealth. These solutions are tested on modern computing platforms to ensure robustness, usability, and scalability. UvA’s research advances the maturity of privacy-by-design AI systems, contributing key innovations that strengthen LICORICE’s goal of trustworthy digital identity and data protection frameworks.
CONTACT
LINK
https://www.uva.nl/en