Visa Defines the Next Era of Commerce: When AI Becomes the Customer

Visa's B2AI Report - conducted with Morning Consult across 2,000 consumers and 512 business decision-makers - delivers the first large-scale empirical dataset on agentic commerce readiness from a major card network. The headline finding is structural: 53% of US businesses would allow AI agents to negotiate prices or terms directly with other AI agents on their behalf. This is not hypothetical adoption intent - it is declared willingness to cede transactional authority to autonomous systems. The consumer data confirms the AXD founding thesis with statistical precision. 58% are comfortable with AI comparing prices, 55% with AI applying discounts, but only 38% with AI completing a purchase - and just 27% with autonomous spending without limits. This is the consumer trust ceiling in quantified form: comfort degrades as autonomy increases. The trust gradient maps directly to the Autonomy Gradient framework. Visa's most consequential data point is institutional: 36% trust bank-backed AI systems versus only 28% for independent AI agents. This eight-point premium validates the AXD position that trust architecture requires institutional anchoring - the payment network becomes the trust infrastructure layer, not just the transaction rail. The generational data - 48% of Gen Z trust payment network-enabled AI versus 20% of Boomers - confirms that the autonomy gradient is not fixed but shifts with cohort experience. Visa's framing of B2AI as a new commerce category - 'an emerging economic model in which AI agents act as active participants in commercial decision-making and execution, while humans remain accountable for intent and outcomes' - is the delegation design principle stated in Visa's institutional language. Frank Cooper III's statement that 'trust becomes the critical infrastructure' is the commercial validation of the AXD founding claim.

What does Visa's B2AI Report reveal about business readiness for agentic commerce?

Visa's Business-to-AI (B2AI) Report, published on 2 April 2026 and conducted with Morning Consult, surveyed 2,000 US consumers and 512 business decision-makers between 29 January and 6 February 2026. The findings represent the first large-scale empirical dataset on agentic commerce readiness from a major card network.

The headline finding is that 53% of US businesses would allow AI agents to negotiate prices or terms directly with other AI agents on their behalf. This is not aspirational intent - it is declared willingness to cede transactional authority to autonomous systems. A further 88% are willing to provide pricing or inventory data to enterprise AI systems, and 77% are already using or piloting AI in their operations.

Visa defines B2AI as 'an emerging economic model in which AI agents act as active participants in commercial decision-making and execution, while humans remain accountable for intent and outcomes.' This framing mirrors the AXD delegation design principle: the human sets the intent, the agent executes within delegated authority, and accountability remains with the delegator.

How does the consumer trust data confirm the AXD founding thesis?

The consumer data provides statistical confirmation of the trust gradient that the AXD Institute has theorised since its founding. Comfort with AI involvement degrades as autonomy increases: 58% are comfortable with AI comparing prices, 55% with AI applying discounts, 38% with AI completing a purchase, and just 27% with AI spending money autonomously without limits.

This is the consumer trust ceiling in quantified form. It maps directly to the Autonomy Gradient framework: as the agent's degree of autonomous authority increases, the proportion of consumers willing to delegate decreases. The gradient is not binary - it is a spectrum that organisations must design for.

Frank Cooper III, Visa's Chief Marketing Officer, stated: 'People are open to AI acting for them, not instead of them.' This captures the delegation design principle in a single sentence. The consumer is willing to delegate specific tasks within defined boundaries, but resists the transfer of open-ended authority.

What is the institutional trust premium and why does it matter?

The most consequential data point for trust architecture is the institutional trust premium. 36% of consumers trust bank-backed AI systems, 35% trust payment network-enabled AI, but only 28% trust independent AI agents. This eight-point gap between institutional and independent agents validates the AXD position that trust architecture requires institutional anchoring.

The implication is structural: the payment network is becoming the trust infrastructure layer, not just the transaction rail. When consumers delegate purchasing authority to an AI agent, they are more willing to do so when a recognised financial institution underwrites the trust relationship. This positions Visa, Mastercard, and the banking system as the trust anchors of agentic commerce.

For organisations designing agentic experiences, this means that the choice of trust anchor is a design decision with measurable impact on adoption. An independent AI shopping agent faces a structural trust deficit that no amount of capability improvement can overcome without institutional backing.

What does the generational data tell us about the future of agentic commerce?

The generational data reveals that the autonomy gradient is not fixed - it shifts with cohort experience. 48% of Gen Z trust payment network-enabled AI systems, compared to only 20% of Boomers. Among Gen Z and Millennials using AI shopping assistants, nearly half report making purchases they would not otherwise have considered due to AI recommendations.

This suggests that the consumer trust ceiling is not a permanent constraint but a generational one. As digital-native cohorts become the dominant consumer demographic, the threshold for delegating purchasing authority to AI agents will rise. Organisations that build trust architecture now are investing in infrastructure that will compound in value as the trust ceiling lifts.

The 40% of Americans who have already made a purchase they normally would not have considered as a result of using an AI agent or tool represents the early evidence of AI-shaped demand. This is not AI assisting existing purchase intent - it is AI generating new purchase intent. The implications for brand strategy, product discovery, and market structure are profound.

How does Visa's B2AI framing relate to AXD practice frameworks?

Visa's B2AI category maps onto several AXD practice frameworks. The delegation design framework governs how authority is transferred from human to agent - Visa's data shows that businesses are ready to delegate but consumers require guardrails. The trust architecture framework provides the structural material for those guardrails - institutional backing, spending limits, and override capability.

The autonomy gradient framework explains the trust decay curve: as autonomy increases, trust decreases unless institutional anchoring compensates. The consumer trust ceiling concept - the maximum level of autonomous authority that a given population segment will accept - is now empirically quantified by Visa's data.

Cooper's statement that 'commerce is moving from market-to-human to market-to-machine' is the commercial articulation of the AXD founding claim: that the design of human-agent relationships requires a new discipline built on trust architecture, delegation design, and outcome specification rather than screen-based interaction.