IEEE Access (Jan 2020)

On the Convergence of Artificial Intelligence and Distributed Ledger Technology: A Scoping Review and Future Research Agenda

  • Konstantin D. Pandl,
  • Scott Thiebes,
  • Manuel Schmidt-Kraepelin,
  • Ali Sunyaev

DOI
https://doi.org/10.1109/ACCESS.2020.2981447
Journal volume & issue
Vol. 8
pp. 57075 – 57095

Abstract

Read online

Developments in artificial intelligence (AI) and distributed ledger technology (DLT) currently lead to lively debates in academia and practice. AI processes data to perform tasks that were previously thought possible only for humans. DLT has the potential to create consensus over data among a group of participants in untrustworthy environments. In recent research, both technologies are used in similar and even the same systems. This can lead to a convergence of AI and DLT, which in the past, has paved the way for major innovations of other information technologies. Previous work highlights several potential benefits of a convergence of AI and DLT but only provides a limited theoretical framework to describe upcoming real-world integration cases of both technologies. In this research, we review and synthesize extant research on integrating AI with DLT and vice versa to rigorously develop a future research agenda on the convergence of both technologies. In terms of integrating AI with DLT, we identified research opportunities in the areas of secure DLT, automated referee and governance, and privacy-preserving personalization. With regard to integrating DLT with AI, we identified future research opportunities in the areas of decentralized computing for AI, secure data sharing and marketplaces, explainable AI, and coordinating devices. In doing so, this research provides a four-fold contribution. First, it is not constrained to blockchain but instead investigates the broader phenomenon of DLT. Second, it considers the reciprocal nature of a convergence of AI and DLT. Third, it bridges the gap between theory and practice by helping researchers active in AI or DLT to overcome current limitations in their field, and practitioners to develop systems along with the convergence of both technologies. Fourth, it demonstrates the feasibility of applying the convergence concept to research on AI and DLT.

Keywords