Governing AI and Blockchain in Finance: A Layered Trust Framework

Divyansh Mishra, R. Agarwal, R. Mishra

Abstract

AI is increasingly embedded into all aspects of financial markets; from credit decisions, compliance processes, trading strategies to payment systems. At the same time, blockchain-based finance has evolved from a speculative niche product into an integral part of the regulated market infrastructure. An integrative review of these two trends is necessary to examine their potential for convergence and how this will necessitate changes in governance. The review uses foundational academic literature in blockchain economics, decentralized finance and algorithmic fairness, as well as regulatory, supervisory and market sources up to October 2026. The authors provide three key findings. Firstly, there has been a structural change in the market: stablecoin capitalizations are estimated to be between approximately USD 270 billion and USD 320 billion. Tokenized real-world assets are concentrated primarily in U.S. Treasuries. The transitional period for the EU Markets in Crypto-Assets Regulation (MiCAR) ends on July 1, 2026. Finally, the U.S. GENIUS Act's rulemaking deadline has passed without a full set of final rules. Secondly, there remains a persistent gap in governance: international reviews show that the implementation of crypto-asset standards varies widely. Furthermore, most security losses result from a few very large attacks. Thirdly, AI agents have started transacting on stablecoin rails prior to the development of supervisory approaches to algorithmic accountability. Finally, the authors propose a five-layered Layered Trust Framework which includes settlement integrity, protocol and code assurance, data and oracle governance, algorithmic accountability, and institutional coordination. Each layer is linked to characteristic risks and governance instruments as well as measurable indicators. Based upon these relationships, the authors derive five testable propositions for future empirical research. The authors conclude that trust in converged AI and blockchain finance is generated by multiple layers of assurance rather than any single technology. Additionally, they believe that the data and algorithmic layers currently represent the weakest areas of governance. Limitations associated with an integrative design, such as relying upon secondary market trackers, are also addressed. Keywords: blockchain; financial technology; AI governance; stablecoins; tokenisation; regulatory architecture

Source: semanticscholar · PDF

Read the AI summary, key takeaways and discussion on WOBR Quant Research.