Assessing the impact of agentic commerce on issuer fraud detection and data integrity

Abstract

This paper shows that the advent of agentic commerce, in which artificial intelligence (AI) agents act as proxies for consumers, represents a structural shift that challenges the trust architecture of digital payments. Although agentic commerce promises major gains in convenience and automation, the paper demonstrates that inserting an AI agent between the consumer and the merchant creates a ‘data chasm’ that interrupts the flow of behavioural, device and contextual data on which modern fraud detection and risk assessment models depend. These data streams allow issuers to infer whether a transaction is legitimate. Without them, the issuer’s view of the transaction environment becomes materially thinner. While emerging agentic commerce protocols propose alternative trust mechanisms, such as cryptographic mandates, agent verification registries and structured delegation frameworks, the analysis shows that these signals are qualitatively different from, and potentially less robust than, the multi-layered behavioural indicators they replace. As a result, issuers are likely to experience greater uncertainty, adopt more conservative risk postures and generate higher rates of false positive declines, reducing merchant authorisation rates. The paper thus identifies a fundamental tension in the shift to agentic commerce, showing how efforts to streamline the front-end experience may ultimately erode the trust mechanisms that have enabled decades of seamless authorisation, destabilising the back-end processes on which modern digital payments depend. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.