The European project DataBri-X, which began in October 2022, has recently concluded, achieving a key milestone: ensuring that data can be shared securely, preventing leaks and unauthorized copies. Within this framework, IMDEA Networks has developed a watermarking tool called FreqyWM that allows institutions to tag their data with a unique, invisible signature.

This advancement will enable business, governmental, public research, and civil society actors to exchange and reuse information, benefiting from its value while complying with legal frameworks on security and privacy. FreqyWM not only was developed but also got integrated with other tools in complex AI pipelines that were demonstrated in real-world use-cases by the DataBri-X consortium.

“This project aimed to demonstrate a no-code, easy-to-use pipeline composition framework for working with European data spaces. The idea is to hide much of the complexity of distributed, privacy-preserving computing behind a simple graphical user interface that allows users to drag and drop existing tools from a marketplace to compose complex services on shared data,” explains Nikolaos Laoutaris, Principal Investigator of DataBri-X at IMDEA Networks.

One of the most significant aspects of DataBri-X is that it provides a holistic, energy-efficient, and user-friendly toolbox of Bri-X “building blocks” (processes, technologies, and tools) that are practical, robust, and scalable. These improve interoperability, usability, data and metadata quality, and integrity, preparing datasets for greater digital value creation within the European Data Spaces context.

As Laoutaris points out, “For European initiatives such as Gaia-X and International Data Spaces (IDSA) to succeed, the technology must be more accessible to non-specialist users. DataBri-X has shown that this is possible. The adoption of these technologies is critical for Europe to remain competitive in the data economy and the development of AI.”

To test the methodology and composition tools, the research team organized two datathons. These events demonstrated that the tools can indeed be used by individuals without specific technical training and can be deployed in a few hours instead of several months.

Moreover, the ability for institutions to watermark their data for the first time provides shared commercial data in AI spaces with protection similar to that already enjoyed by movies or audio tracks on platforms such as Netflix or Spotify. This research was published at ICDE 2024 under the title “FreqyWM: Frequency Watermarking for the New Data Economy” (D. Isler, E. Cabana, A. Garcia-Recuero, G. Koutrika, N. Laoutaris).

DataBri-X exemplifies how innovative and reliable tools can be developed to facilitate secure information exchange in the emerging European data economy, promoting trustworthy AI aligned with European values.