The payments industry has spent decades engineering trust between human buyers and merchants — building fraud detection systems, authentication layers, and dispute mechanisms calibrated to the rhythms of human decision-making. Now, a fundamental disruption is underway: artificial intelligence agents are beginning to transact autonomously on behalf of consumers and businesses, and the existing trust architecture was never designed for them. It is within this context that J.P. Morgan Payments is making one of its most consequential strategic bets, placing Prashant Sharma — Executive Director of Biometric Payments and Agentic Commerce for Merchant Services — at the centre of an effort to construct the governance and trust frameworks that this new commercial era will demand.
Sharma's mandate is, in its own way, a signal of how seriously the largest bank in the United States is treating the agentic commerce opportunity. His dual remit — spanning both biometric payments and agentic commerce within the Merchant Services division — reflects J.P. Morgan's recognition that authentication and autonomous agency are two sides of the same coin. If an artificial intelligence agent is to conduct a transaction on a user's behalf, the payments rail beneath it must be capable of verifying not just that a credential is valid, but that the agent acting on that credential is authorised, trustworthy, and operating within sanctioned parameters. That is an entirely different engineering and governance challenge from the one payments networks solved in the era of card-present transactions.
From Strategy to Scaffolding
The language Sharma uses is precise and instructive. J.P. Morgan Payments' primary objective, as he has articulated it, is to translate agentic commerce ambitions into practical scaffolding — governance structures and trust frameworks that can actually be operationalised at scale. This distinction between strategy and scaffolding matters enormously. The financial services industry has generated substantial enthusiasm around agentic artificial intelligence, with institutions announcing pilots, publishing white papers, and signalling transformation. What has lagged behind is the harder, less glamorous work: defining who is liable when an agent makes an erroneous purchase, how merchants verify the legitimacy of an agent-initiated transaction, and what standards govern the delegation of spending authority from a human principal to a machine intermediary.
J.P. Morgan's positioning is deliberately infrastructure-first. Rather than leading with consumer-facing products, Sharma's team appears focused on building the underlying architecture — the rails of accountability, the verification protocols, and the governance models — that would allow agentic commerce to function reliably across the merchant ecosystem. This is consistent with J.P. Morgan's broader institutional instinct: the bank has historically preferred to own the plumbing rather than compete at the application layer, a posture that has served it well in treasury services, wholesale payments, and cross-border transaction infrastructure.
Why Biometrics and Agentic Commerce Belong Together
The pairing of biometric payments with agentic commerce within a single executive portfolio is not incidental. Biometric authentication — encompassing fingerprint, facial recognition, and behavioural signals — represents the most robust method currently available for binding a transaction to a verified human identity. As commerce increasingly flows through AI intermediaries, that binding becomes more complex and, arguably, more critical. An agent acting on behalf of a cardholder must somehow inherit verified authority without creating a vector for fraud or unauthorised action. Sharma's team sits at precisely that intersection, tasked with designing systems where biometric-grade trust can be extended into, or appropriately adapted for, agent-driven environments.
This is not merely a technical problem. It is a regulatory and liability question that no jurisdiction has yet fully answered. Who bears responsibility when an AI agent, properly credentialed, executes a transaction that the human principal later disputes? How do merchants adjust their chargeback and dispute infrastructure to accommodate agent-originated orders? These are the questions that a credible trust framework must address before agentic commerce can move from pilot to mainstream. J.P. Morgan's willingness to invest institutional capital — human, financial, and reputational — in answering them positions the bank as a potential standard-setter rather than a fast follower.
What This Means for the Industry
The implications of J.P. Morgan Payments' agentic commerce initiative extend well beyond the bank's own merchant relationships. If Sharma's team succeeds in developing a credible governance framework, that framework could become a de facto industry benchmark — much as J.P. Morgan's treasury and risk management practices have shaped corporate finance conventions for generations. Merchants, payment service providers, and technology platforms building agentic capabilities will all eventually require a trust layer that can interoperate with the incumbent payments infrastructure. A bank of J.P. Morgan's scale that moves first on that architecture gains considerable leverage in defining what interoperability looks like.
For the broader fintech ecosystem, the message is equally significant. The race to build agentic commerce is not simply a race to deploy large language models or automate purchasing workflows. It is, at its foundation, a race to establish trust — and trust in payments has always been an institutional product, built on regulatory compliance, liability frameworks, and decades of counterparty relationships. J.P. Morgan Payments, through Prashant Sharma's work, is asserting that those institutional assets are not legacy constraints but competitive advantages in the agentic era. Whether that assertion proves correct will depend on execution, but the strategic logic is coherent and the institutional commitment appears genuine.
Written by the editorial team — independent journalism powered by Codego Press.