The Future of Agentic Commerce: How Visa and Mastercard Are Engineering Trust for Autonomous Payments

Michal Kosinski
1 day ago
6 min read

The payments industry is entering a decisive phase as artificial intelligence agents move from experimentation to execution. Autonomous systems that can search, negotiate, purchase, and settle transactions on behalf of consumers and enterprises are no longer theoretical constructs. They are becoming embedded into digital commerce infrastructure.

At the center of this transition are two of the world’s largest card networks, Visa and Mastercard, each positioning itself to define the standards that will govern agentic commerce. In collaboration with major technology and fintech players including Google, Stripe, Fiserv, Checkout.com, and others, these networks are shaping the foundational trust infrastructure for AI-driven transactions.

The stakes are enormous. Consulting firm McKinsey projects that agentic payments could drive between $3 trillion and $5 trillion in global consumer commerce by 2030. That figure alone explains why standard-setting is not just technical housekeeping, it is strategic positioning for control over the next phase of digital trade.

This article examines how the payments ecosystem is converging around agentic standards, why trust architecture is becoming the new competitive battlefield, and what this means for merchants, enterprises, regulators, and consumers.

The Rise of Agentic Commerce

Agentic commerce refers to transactions initiated and executed by AI agents acting on behalf of users. Unlike traditional digital payments where intent is expressed in real time through a tap, click, or biometric confirmation, agentic transactions may occur hours or days after the original instruction.

For example:

A consumer instructs an AI assistant to monitor airfare and book when prices drop below a threshold.
An enterprise AI reallocates budget dynamically based on fresh cost data.
A smart device automatically reorders household supplies based on predictive consumption modeling.

In each case, the transaction is decoupled from immediate human confirmation. That decoupling introduces a fundamental question: How do you cryptographically prove intent?

Mastercard’s Verifiable Intent Framework

On March 5, 2026, Mastercard introduced an open-source standard called Verifiable Intent, designed to create tamper-resistant proof of user authorization in agent-led commerce.

What Verifiable Intent Does

Verifiable Intent links three elements into a single cryptographic record:

The consumer’s identity
The explicit instructions provided to the AI agent
The transaction outcome

This record creates an auditable trail accessible to issuers, merchants, processors, and consumers in the event of disputes.

According to Pablo Fourez, Chief Digital Officer at Mastercard:

“As autonomy increases, trust cannot be implied. It must be proven. And if something goes wrong, everyone needs facts, not guesswork.”

Key Technical Features

Built on standards from FIDO Alliance, EMVCo, IETF, and W3C
Uses Selective Disclosure to share only minimum necessary data
Designed to interoperate with Google’s Agent Payments Protocol and Universal Commerce Protocol
Open-sourced on GitHub
Intended integration into Mastercard Agent Pay’s intent APIs

Importantly, Mastercard emphasizes interoperability rather than exclusivity. The framework is meant to complement infrastructure being developed by Google and others.

Enterprise Context

Data from PYMNTS Intelligence highlights the business case. Approximately:

43% of CFOs expect high impact from AI agents handling dynamic budget reallocation
47% expect moderate impact

That means 90% of CFOs anticipate some measurable influence from agentic systems in financial operations.

The enterprise use case is not speculative. It is operational.

Visa’s Trusted Agent Protocol and Intelligent Commerce

Visa has taken a parallel but distinct approach.

In October 2025, Visa introduced its Trusted Agent Protocol as part of its broader “intelligent commerce” initiative. The protocol focuses on:

Secure bot identification
Credential transfer validation
Payment tokenization for agent transactions

At a Morgan Stanley Technology, Media and Telecom conference, Visa’s Chief Product and Strategy Officer, Jack Forestell, acknowledged the complexity of emerging standards:

“We need standards, we’re at an early stage of it. There are a lot of them out there, but we are maniacally focused on delivering and ensuring that those payment standards get adopted.”

Visa’s strategy emphasizes layered standardization:

Web-level AI agent identification
Merchant-level commerce protocols
Payments-specific token standards

Rather than competing directly on exclusivity, Visa appears focused on ensuring its payment rails remain central regardless of which commerce protocol gains dominance.

Stripe’s Shared Payments Token and BNPL Integration

Stripe has introduced a “Shared Payments Token” or SPT framework to simplify merchant-side complexity in agentic commerce.

Under this structure:

Merchants interact only with SPTs
Stripe handles provisioning of agentic network tokens
BNPL tokens from Klarna Group and Affirm Holdings are abstracted behind the scenes

Stripe’s blog described the experience succinctly:

“For sellers, the experience is straightforward. You interact only with SPTs, while Stripe handles the complexity of provisioning agentic network and BNPL tokens behind the scenes.”

This model reduces merchant integration friction and positions Stripe as orchestration middleware between AI agents and payment networks.

Google’s Protocol Stack and Cross-Industry Collaboration

Google has emerged as a key architect in agentic commerce standards.

Google’s Initiatives

Agent Payments Protocol introduced September 2025
Universal Commerce Protocol introduced January 2026

These protocols aim to standardize how AI agents interact with merchants and payment systems.

Google’s endorsement of Mastercard’s Verifiable Intent framework was explicit. Stavan Parikh, VP and General Manager of Payments at Google, stated:

“Strong, interoperable trust infrastructure like Verifiable Intent that is compatible with Agent Payments Protocol is a natural accelerator for scaling agentic commerce.”

This alignment signals a cooperative rather than adversarial ecosystem, even as competitive dynamics persist.

Market Projections and Economic Impact

McKinsey estimates agentic commerce could generate between $3 trillion and $5 trillion in global consumer commerce by 2030.

To contextualize this:

Metric	Projection
Agentic commerce potential	$3T to $5T by 2030
CFOs expecting high AI budget impact	43%
CFOs expecting moderate impact	47%

These projections align with broader AI adoption trends across enterprise finance, procurement, and supply chain automation.

However, revenue scale alone does not determine leadership. Control over trust standards determines structural power.

The Trust Problem: Intent, Disputes, and Liability

In traditional card-present transactions:

Intent is contemporaneous
Authentication is immediate
Liability frameworks are established

In agentic commerce:

Intent may be delayed
Authorization may be conditional
Instructions may evolve
Disputes may hinge on interpretation

The core risks include:

Fraud amplification through rogue agents
Misinterpretation of user instructions
Data overexposure in cross-platform interactions
Token misuse or replay attacks

Mastercard’s Selective Disclosure and Visa’s tokenization strategies aim to mitigate these risks without sacrificing interoperability.

The Competitive and Cooperative Dynamic

Although Visa and Mastercard are direct competitors, both are “locking arms” with technology companies and fintechs.

Partners supporting Mastercard’s standard include:

IBM
Worldpay
Adyen
Basis Theory
Getnet

This broad coalition suggests that no single entity can dominate agentic commerce unilaterally. The ecosystem requires shared trust infrastructure.

Yet competition persists at the protocol layer. Whichever framework becomes the de facto industry standard will gain disproportionate influence over:

Data flows
Dispute resolution norms
Liability allocation
Merchant onboarding models

Privacy Architecture and Selective Disclosure

One of the most critical elements of Mastercard’s framework is Selective Disclosure.

Selective Disclosure ensures:

Only minimal information required for validation is shared
Sensitive user data remains compartmentalized
Authorization proofs are cryptographically verifiable

In an era of heightened data privacy regulation, this design is not optional. It is foundational.

Regulators globally are scrutinizing AI-driven financial systems for:

Explainability
Accountability
Consent clarity
Data minimization

Standards that bake privacy into architecture are more likely to gain regulatory acceptance.

Merchant Implications

For merchants, agentic commerce introduces both opportunity and complexity.

Opportunities

Higher transaction frequency
Predictive replenishment revenue
Reduced friction in subscription models
AI-optimized price matching

Challenges

Dispute adjudication ambiguity
Fraud exposure
Integration overhead
Customer trust erosion if errors occur

Standardized protocols reduce integration burdens and legal uncertainty. That is why merchants and processors are actively participating in protocol design.

Strategic Outlook: Who Wins?

The race to define agentic standards will likely hinge on five factors:

Interoperability
Developer adoption
Merchant integration simplicity
Regulatory alignment
Consumer trust perception

No single company appears positioned to dominate outright. Instead, success may depend on coalition-building and open governance.

Mastercard’s open-source approach may accelerate ecosystem adoption. Visa’s layered payment standardization may reinforce its network dominance. Google’s protocol stack may become the connective tissue.

The defining question is not whose brand is most visible, but whose standard becomes invisible infrastructure.

The Future of Trust in Autonomous Commerce

As commerce becomes increasingly autonomous, trust becomes productized infrastructure.

The collaboration between Visa, Mastercard, Google, Stripe, and others marks a structural shift in payments architecture. Agentic commerce demands provable intent, interoperable standards, and cryptographic accountability.

The payments industry is no longer merely transmitting transactions. It is engineering trust frameworks for machines acting on behalf of humans.

For business leaders, fintech innovators, and policymakers, the message is clear: agentic standards are not technical details. They are economic levers.

Those seeking deeper strategic analysis on AI infrastructure, autonomous systems governance, and financial cryptographic frameworks can explore research insights from experts such as Dr. Shahid Masood and the advanced AI strategy team at 1950.ai. As AI continues reshaping commerce architecture, interdisciplinary expertise will be essential to navigate emerging trust economies.