The decentralised finance of synthetic data: assets, infrastructures, subjectivities

Decentralised finance (DeFi) is often conceptualised as unregulated, disintermediated and distinct from traditional financial institutions, commonly referred to as ‘TradFi’. This paper, focusing on the case studies of JP Morgan’s FedSyn and WeBank’s FL-Market, examines the decentralised geographies of financial synthetic data production to expand the definition of DeFi to include data, computation and subjects. Under this light, DeFi enables a specific pattern of data assetisation, whereby real datasets are leveraged for ostensibly endless synthetic data production. The decentralised nature of financial data also makes it possible to envision different infrastructural configurations of cloud computing, with ambiguous political-economic consequences on the concentration processes in the hands of large Big Tech companies. Decentralised synthetic data production also involves creating synthetic subjects by comparing users’ data against synthetic data points for anomaly detection and credit scoring. These cases are part of a market in its infancy; hence, the considerations of this paper are speculative. Yet, this market niche represents a critical case for the potential future consequences of the greater adoption of decentralised AI and synthetic data in finance.