No abstract

One of the foundations of data sharing in the European Union (EU) is trust, especially in view of the advancing digitalization and recent developments with respect to European Data Spaces. In this chapter, we argue that privacy-preserving techniques, such as multi-party computation and fully homomorphic encryption, can play a positive role in enhancing trust in data sharing transactions. We therefore focus on an interdisciplinary perspective on how privacy-preserving techniques can facilitate trustworthy data sharing. We start with introducing the legal landscape of data sharing in the EU. Then, we discuss the different functions of third-party intermediaries, namely, data marketplaces. Before giving a legal perspective on privacy-preserving techniques for enhancing trust in data sharing, we briefly touch upon the Data Governance Act (DGA) proposal with relation to trust and its intersection with the General Data Protection Regulation (GDPR). We continue with an overview on the technical aspects of privacy-preserving methods in the later part, where we focus on methods based on cryptography (such as homomorphic encryption, multi-party computation, private set intersection) and link them to smart contracts. We discuss the main principles behind these methods and highlight the open challenges with respect to privacy, performance bottlenecks, and a more widespread application of privacy-preserving analytics. Finally, we suggest directions for future research by highlighting that the mutual understanding of legal frameworks and technical capabilities will form an essential building block of sustainable and secure data sharing in the future

Big data has been a pervasive catchphrase in recent years, but dealing with data scarcity has become a crucial question for many real-world deep learning (DL) applications. A popular methodology to efficiently enable the training of DL models to perform tasks in scenarios where only a small dataset is available is transfer learning (TL). TL allows knowledge transfer from a general domain to a specific target one; however, such a knowledge transfer may put privacy at risk when it comes to sensitive or private data. With CryptoTL we introduce a solution to this problem, and show for the first time a cryptographic privacy-preserving TL approach based on homomorphic encryption that is efficient and feasible for real-world use cases. We demonstrate this by focusing on classification tasks with small datasets and show the applicability of our approach for sentiment analysis. Additionally we highlight how our approach can be combined with differential privacy to further increase the security guarantees. Our extensive benchmarks show that using CryptoTL leads to high accuracy while still having practical fine-tuning and classification runtimes despite using homomorphic encryption. Concretely, one forward-pass through the encrypted layers of our setup takes roughly 1 s on a notebook CPU.Preprint. Under review.