The study sketches a framework towards the development of a European AI ecosystem underpinned by a sovereign data marketplace and federated learning. The approach presented aims to strengthen Europe's technological sovereignty in the age of Artificial Intelligence (AI) by outlining a blockchain-based vision for ‘AI Made in Europe’. The focus is on the integration of blockchain technology and AI in order to optimize data processing and create new opportunities to efficiently manage the growing volumes of data. A key potential of blockchain lies in its ability to enable fairer access to data and to productively synchronize contributions to the network. The system builds on the existing decentralized data infrastructure in Europe and sees the federated approach as an opportunity to understand European diversity as a global competitive advantage in the development of AI models.

The approach is based on three main components:
1.    Digitalization and data infrastructures: the synergies of AI and blockchain enable data protection-compliant collaboration, which is business-critical given the increasing digitalization of the economy and rising regulatory requirements for data management in the EU. 
2.    Companies: Start-ups and small and medium-sized enterprises (SMEs) in particular benefit from AI blockchain synergies. Due to high market barriers and data protection costs, they have so far been less able to participate in the economic potential of AI compared to larger corporations.
3.    Data marketplaces: Decentralized digital infrastructures can be created through the combined use of blockchain and AI. These enable individuals, start-ups, and companies to jointly train AI models and use their results or send algorithms for data processing without disclosing the content of their data. Such marketplaces can be used by both individual and corporate players and facilitate access to valuable data sources.

Fields of application: Potential areas of application for digital data marketplaces based on blockchain and AI include healthcare, the development of smart cities, industrial AI applications to optimize machine production processes and AI training on mobile devices. Other conceivable areas include the metaverse and the traceability of data to verify compliance with AI regulations.

You can read the entire study as a PDF here.