In the letter, part of the company’s annual report, he described a “rewiring” of commerce driven by agentic artificial intelligence (AI), tokenized credentials and data, stablecoins and digitized identity. He argued that every new way to pay (from account-to-account and wallets to crypto and “agent-led commerce”) only matters if it can scale with security, reliability and modern risk controls.
McInerney put agentic AI at the front of the platform shift. In his view, the next wave of digital shopping won’t start with a consumer typing, tapping or even browsing. It will start with an intelligent agent. He noted the breakneck pace of consumer AI adoption and says AI-driven referral traffic is already reshaping how shoppers find products and how retailers capture demand.
But as those agents enter the payment flow, Visa sees a familiar problem in a new disguise: trust. If the interface to commerce becomes a bot, the system still has to know whose money is being spent, what permissions the agent has, and how credentials are presented and authenticated without opening the door to new fraud and cyber risks. McInerney pointed to Visa Intelligent Commerce, launched in April 2025, as Visa’s framework to provide agents with Visa credentials while enabling trust, security and interoperability across the ecosystem. He also highlighted the Visa Trusted Agent Protocol, designed to let sellers use existing websites as agent-friendly storefronts, speeding time to market and simplifying integration.
“Generative AI-driven, or agentic, commerce is transforming the digital shopping experience,” he stated in the letter. “This shift will rival the impact that online shopping and mobile devices have had on commerce in past decades.”
Beyond AI, McInerney’s message is that Visa is trying to make “trust” programmable by packaging network capabilities into modular services that partners can consume through integrations, APIs and structured data exchange. The letter calls out several technologies shaping how consumers pay and how money moves:
Advertisement: Scroll to Continue
- Tap-to-everything: Tap has moved from a “nice to have” to the default in-person gesture, and Visa is expanding it well beyond the terminal. Tap to Phone turns smartphones into acceptance devices for small sellers, while Tap to Add Card and newer “tap” interactions (including activation and confirmation flows) aim to streamline onboarding and verification.
- Tokenization: McInerney argued the era of manual key-entry is fading as card-on-file and wallet experiences replace friction-heavy checkout. Tokenization replaces sensitive card data with a unique digital identifier that can only be used in the context it was issued and Visa is augmenting tokens with enhanced data and passkeys to improve both security and conversion.
- Risk and security: As commerce gets faster, attackers adapt. McInerney pointed to AI-powered tools such as Visa Protect for A2A for fraud detection and prevention, and emphasizes scam disruption efforts with clients and law enforcement. He also notes Visa is extending these defenses into agent-based commerce tools as assistants become a new attack surface.
- Stablecoins as next-gen settlement infrastructure: He portrayed stablecoins as early-stage but potentially foundational, because they can unify messaging and settlement on shared ledgers that operate continuously. Visa’s roadmap includes stablecoin-linked cards, settlement in stablecoins such as USDC, stablecoin prefunding for cross-border payouts and pilots that deliver payouts directly to stablecoin wallets.
- “Visa as a Service”: McInerney described a layered stack with global connectivity at the base, modular services (credentials, tokens, authentication, risk management) above it, solutions assembled from those components, and an access layer delivered via APIs and custom integrations. The aim is to let banks and fintechs plug into Visa’s capabilities quickly, without rebuilding the trust layer from scratch.
“The convergence of AI, tokenization, digital identity, stablecoins and more is not simply upgrading the way money moves,” he wrote. “It is creating a network that thinks, learns and protects in real-time.”
