English

AI Agent Economy: Visa, Mastercard Vie for Payment Protocol Dominance

The Fundamental

Nowpattern

10 5月 2026 — 11 min read

⚡ FAST READ

AI agents are fundamentally changing the structure of commerce, and companies controlling payments, automation, and foundational models could establish new platform dominance.

PATTERN: Platform Dominance × Winner-Take-All

CORE SCENARIO: Companies that dominate the infrastructure of the AI agent economy will collect platform taxes across payment, automation, and inference layers, rewriting the rules of commerce.

FOCUS: How Visa and Mastercard will establish standard protocols for AI agent payments and compete for market share.

Why it matters: The same week an experiment where an AI agent was given $50 to buy a PC became a hot topic, Visa predicted that "millions of people will shop via AI agents by the 2026 holiday season." Meanwhile, Gartner announced that "by 2028, AI agents will mediate 90% of B2B purchases, reaching a scale of $15 trillion." This is not merely a tech experiment. It is a turning point where the fundamental structure of commerce – "who buys, who sells, and who pays" – is being rewritten.

GENRE:#TECHNOLOGY #ECONOMY_TRADE

EVENT:#TECHNOLOGICAL_ADVANCEMENT

DYNAMICS (Nowpattern):#PLATFORM_DOMINANCE #LEAPFROGGING

📝 SUMMARY: The same week an experiment where an AI agent was given $50 to buy a PC became a hot topic, Visa predicted that "millions of people will shop via AI agents by the 2026 holiday season."

What Happened

February 2026 — A Reddit user published an experiment giving an AI agent $50 to autonomously attempt a PC purchase. The agent automatically performed price comparison, spec evaluation, and purchase decisions.
February 2026 — A fully automated reservation workflow for a restaurant AI voice agent was published on the n8n platform. It operates 24/7 unmanned with VAPI + PostgreSQL + Twilio integration.
November 2025 — Gartner predicted that "by 2028, AI agents will mediate 90% of B2B purchases, reaching a scale of $15 trillion" (Gartner IT Symposium/Xpo 2025).
December 2025 — Visa predicted that "millions of people will purchase via AI agents by the 2026 holiday season." Announced the Intelligent Commerce program.
October 2025 — Visa and Mastercard simultaneously announced agentic payment tools. Mastercard introduced "Agent Pay," and Visa introduced "Trusted Agent Protocol."
October 2025 — n8n raised $180M (led by Accel, with NVidia NVentures participation), reaching a valuation of $2.5B. ARR reached $40M.
January 2026 — The World Economic Forum announced that "the AI agent market will reach $236B by 2034."
2025 — Retail site traffic via AI agents on Black Friday increased by 805% year-over-year. AI-driven sales exceeded $22B worldwide.
2025 — Coinbase announced Agentic Wallets, enabling AI agents to autonomously hold, send, and trade crypto assets.
February 2026 — Cloudflare announced a security framework for agentic commerce, an AI agent authentication infrastructure compatible with Visa and Mastercard.

Overall Picture

Historical Context

For the past 30 years, e-commerce has been built on the premise that "humans shop in browsers." Amazon (1994), Rakuten (1997), Shopify (2006) — all platforms fundamentally assume "human clicks."

This premise began to crumble from late 2025. There are three triggers.

First, the dramatic improvement in LLM (Large Language Model) "tool-use capabilities." GPT-4o, Claude 3.5, and Gemini 2.0 gained the ability to call APIs, operate the Web, and input payment information.

Second, the adaptation of payment infrastructure. With Visa and Mastercard simultaneously announcing agentic payment tools in October 2025, the institutional foundation for "AI to pay" was established.

Third, the maturation of no-code automation platforms. n8n achieved a $180M funding round and a $2.5B valuation in 2025, providing an environment where even non-engineers can build and deploy AI agents.

These three elements converged simultaneously from late 2025 to early 2026. Historically, this marks the beginning of a structural transformation comparable to "the mid-1990s when the internet entered commerce."

When PizzaHut launched online ordering in 1994, many laughed, saying "ordering food on a PC is niche." Two years later, Amazon began selling books, and five years later, e-commerce transformed the core of commerce. The same pattern is now beginning as a shift from "humans buying" to "AI buying."

Stakeholder Map

ACTOR	STATED GOAL	TRUE MOTIVE	✅ GAINS	❌ LOSSES
Visa / Mastercard	Improved payment security	Desire to monopolize the payment layer of the agent economy	Status as standard protocol setter for AI agent payments	Traditional fee models compressed by low margins of machine-to-machine transactions
n8n / No-code Platforms	Democratization of automation	Desire to earn platform tax as the "infrastructure layer" of the AI agent economy	Corporate reliance on AI automation (ARR $40M, valuation $2.5B)	Limits of open-source commercialization, forks by AWS and Google
OpenAI / Anthropic / Google	Safe development of AI	Desire to dominate the OS (foundational model) of the agent economy	Explosive increase in API calls and pay-as-you-go revenue	Increased regulation due to agent runaway incidents
Amazon / Retail Platforms	Improved consumer experience	Desire to be the default purchase destination for AI agents	Capturing sales via agents	Constant exposure to price comparison AI, loss of brand loyalty
Consumers	Convenient and affordable shopping	Desire to save time and effort	Price optimization, time saving	Black-boxing of purchase decisions, loss of privacy
SMEs (Restaurants, etc.)	Operational efficiency	Reduced labor costs and 24/7 availability	Automated reservations with n8n voice agent, 24/7 availability for $50/month server cost	Risk of elimination by becoming an "invisible store" if not AI agent compatible

Structure Seen Through Data

$15 Trillion — Gartner forecast: Total B2B purchase value mediated by AI agents in 2028
90% — Gartner forecast: Proportion of B2B purchases mediated by AI agents in 2028
$236B — World Economic Forum forecast: AI agent market size in 2034
805% — Year-over-year increase in retail traffic via AI agents on Black Friday 2025
$22B+ — Worldwide sales generated by AI on Black Friday 2025
$2.5B — n8n's valuation as of October 2025 ($180M raised, led by Accel)
$40M ARR — n8n's Annual Recurring Revenue (as of July 2025)
36% — Percentage of consumers who responded they are open to AI agents making purchases on their behalf
$50/month — Server cost for 150,000 executions per month for n8n self-hosted

The delta: Superficially, it appears to be a tech experiment where "AI shops," but at its core, it's a civilizational shift where "the human role in commerce changes." Visa and Mastercard's simultaneous entry into agent payments is because they are convinced that "the next customer will be an algorithm, not a human."

Reading Between the Lines — What the News Isn't Saying

While Visa and Mastercard outwardly tout improved security for AI agent payments, their true motive is to monopolize the payment layer of the agent economy and maintain fee revenue. However, with fees compressed in machine-to-machine transactions, they are compelled to seek new revenue streams. Foundational model companies like OpenAI also aim for increased API calls and pay-as-you-go revenue under the guise of safe AI development, but they fear increased regulation due to the risk of agent runaway.

NOW PATTERN

Platform Dominance × Winner-Take-All

Those who control the infrastructure layer of the AI agent economy (payments, automation, foundational models) will write the rules of commerce for the next 30 years.

Platform Dominance: The War for "Who Controls the Agent's Wallet"

In October 2025, Visa and Mastercard announced agentic payment tools in the same week. This was no coincidence. Both companies have built empires over 30 years on the act of "humans presenting cards." They moved simultaneously precisely because they saw a future where that premise disappears.

Visa's Intelligent Commerce program gives developers APIs that plug identity checks, spending controls, and tokenized card credentials into AI agents.
— Visa Investor Relations, December 2025

Mastercard's Agent Pay uses 'agentic tokens' and requires that AI agents be registered and verified before they can initiate payments.
— Mastercard Press Release, October 2025

Visa's "Trusted Agent Protocol" and Mastercard's "Agent Pay" are technically mechanisms to connect AI agents to payment networks, but strategically, they represent a battle for supremacy over "the standard for the next 30 years of the payment layer."

To summarize the structure: - Layer 1 (Foundational Models): OpenAI / Anthropic / Google — The "brain" of AI - Layer 2 (Automation Infrastructure): n8n / Make / Zapier — The "limbs" of AI - Layer 3 (Payment Infrastructure): Visa / Mastercard — The "wallet" of AI - Layer 4 (Crypto Asset Payments): Coinbase Agentic Wallets — AI's "other wallet"

The winner of each layer will collect "platform tax" from the agent economy. This is a recreation of the battle fought by Amazon, PayPal, and VISA in the early days of e-commerce in the 1990s, but this time with the fundamental difference that "the customer is an algorithm, not a human."

Unlike humans, AI agents have no brand loyalty. They don't buy based on emotion. They compare all options in 0.1 seconds. This signifies "perfect efficiency in price discovery" and could lead to the structural collapse of brand premiums and retail margins.

Winner-Take-All: The $50 Experiment Reveals the Advent of an "AI-Native Economy"

The "experiment of giving an AI agent $50 to buy a PC" that became a hot topic on Reddit is still technically crude. However, what this experiment demonstrates is not the maturity of the technology, but a cultural shift where "humans have begun to cross the psychological threshold of entrusting shopping decisions to AI."

36% of customers are open to AI agents making purchases on their behalf.
— World Economic Forum, January 2026

By 2028, 90% of all B2B purchases will be handled by AI agents, channeling more than $15 trillion in spending through automated exchanges.
— Gartner IT Symposium/Xpo 2025

While 36% of consumers responded that they are open to AI handling their shopping, this figure closely resembles the early days of e-commerce. In 1998, approximately 30% of consumers said they were willing to enter their credit card numbers online. Five years later, that number exceeded 80%, and e-commerce became irreversibly established.

The essence of Gartner's forecast is to see a similar "crossing of the trust threshold" occurring between 2025 and 2028. Especially in the B2B sector, 90% of recurring orders (MRO: Maintenance, Repair, and Operations parts) by procurement managers can be automated, and the entry of AI agents here faces low technical and organizational barriers.

Meanwhile, the case of the restaurant AI voice agent demonstrates the "winner-take-all" dynamic. An automated reservation system built with n8n + VAPI + PostgreSQL operates 24/7 for a server cost of $50 per month. This 24-hour availability with zero labor costs gives adopting establishments an overwhelming competitive advantage. Conversely, non-adopting establishments will be eliminated as "stores that cannot be reached by phone."

This is a recurrence of the digital divide for SMEs, but this time the divide is more brutal. While the "having a website or not" in the 2000s led to gradual elimination, adapting to the AI agent economy involves a more direct exclusion mechanism: "whether to enter a store recommended by an AI agent or not."

Intersection of Dynamics

The two dynamics — platform dominance and winner-take-all — are inseparably linked in the AI agent economy.

When the platform layer (Visa/Mastercard/n8n/OpenAI) dominates the infrastructure for payments, automation, and inference, only companies on that infrastructure will be among the AI agent's "options." Companies not on the infrastructure will literally become "invisible to AI."

This resembles the structure of SEO (Search Engine Optimization) 20 years ago, but with a fundamental difference. In SEO, humans were the final decision-makers. In the AI agent economy, AI performs recommendations, decisions, and purchases. Humans only hold the intention to "shop," entrusting the entire remaining process to AI.

If Gartner's forecast of "90% B2B purchase automation by 2028" materializes, the structure of global commerce will irreversibly shift from an "economy chosen by humans" to an "economy chosen by algorithms." In this transition, those who control the infrastructure layer will reap 30 years' worth of platform tax.

History of Patterns

1994: The Birth of E-commerce — From PizzaHut Online Ordering to Amazon

In 1994, PizzaHut began taking pizza orders via the internet. In the same year, Jeff Bezos founded Amazon. By 1998, PayPal began providing payment infrastructure.

At the time, consumer trust in "shopping on the internet" was around 30%. However, within five years, it exceeded 80%, and e-commerce became irreversibly established. Crucially, companies that controlled the three layers of "payment infrastructure," "marketplaces," and "logistics" during this transition became the winners for the next 30 years: Visa/PayPal (payments), Amazon (marketplace), FedEx/UPS (logistics).

Structural similarities with the current situation: Emergence period of a new purchasing channel (browser → AI agent). Consumer trust level 30-36%. Battle for payment infrastructure.

2007: iPhone and the Structural Transformation of Mobile Commerce

With the advent of the iPhone in 2007, the main battlefield for e-commerce shifted from desktop to mobile. Apple Pay (2014) established the standard for mobile payments, and the App Store imposed a 30% platform tax.

While "shopping on smartphones" was initially niche, by 2020, over 70% of all e-commerce was conducted via mobile. Again, the infrastructure layer (Apple/Google) captured platform taxes, and non-compliant retailers were eliminated.

Structural similarities with the current situation: Generational change in purchasing interface. Infrastructure layer (Apple → Visa/n8n) captures platform tax. Elimination of non-compliant businesses.

Patterns Revealed by History

The pattern revealed by history is clear. Every time the "interface" of commerce changes, three things happen:

① Initial consumer trust is around 30%, but exceeds 80% within 5 years.

② Companies that control the infrastructure layer collect platform taxes for 30 years.

③ Businesses not compatible with the new interface are eliminated within 5-10 years.

The AI agent economy of 2026 is entering the third cycle of this historical pattern: "buying with a browser" → "buying with a smartphone" → "AI buying." In each cycle, human involvement in decision-making decreases, and the scope of automation expands.

Future Outlook

Base Case — Gradual Penetration — B2B First, C-to-C Mainstream by 2028 (Probability: 55-65%)

Automation of MRO (Maintenance, Repair, and Operations) procurement will lead in the B2B sector. By 2028, 50-70% of B2B purchases will be via AI agents. For C-to-C (consumer-to-consumer), Visa/Mastercard pilots will begin by late 2026, with full adoption by 2028. The adoption rate of AI voice agents by SMEs will exceed 20% in the restaurant industry by late 2027. No-code platforms like n8n will become the standard for "AI agent building for non-engineers."

Implications for Investment/Action: Focus on automation platform-related stocks such as n8n / Make / Zapier (n8n is not publicly traded). Visa / Mastercard will benefit in the medium term as first movers in agentic payments. Automation SaaS for the food service and retail industries (AI voice reservations, automated ordering) will see investment opportunities in 2026-2027.

Optimistic Case — Explosive Growth of the Agent Economy — "AI-Native Commerce" Becomes Mainstream in 2027 (Probability: 15-25%)

OpenAI / Anthropic models become fully autonomous in web browsing and payments. Visa/Mastercard payment tokens are standardized, and sales via AI agents exceed 20% of total e-commerce by the 2027 holiday season. Gartner's $15T forecast begins to materialize ahead of schedule.

Implications for Investment/Action: Strengthen investments in foundational model companies (Microsoft, Google, Amazon, affiliated with OpenAI). Infrastructure companies for agent payments (Cloudflare, Stripe) experience rapid growth. Traditional advertising and marketing companies are structurally hit by the "AI doesn't see ads" problem.

Pessimistic Case — Regulatory Hurdles and Security Incidents — 2-3 Year Delay in Agent Economy (Probability: 15-25%)

Large-scale erroneous orders or fraud incidents by AI agents occur, leading to strengthened consumer protection regulations. Strict enforcement of the EU AI Act mandates additional authentication and approval processes for agent payments. Consumer distrust in AI grows, and purchases via agents remain below 5% of the total even by 2028.

Implications for Investment/Action: Cybersecurity companies (CrowdStrike, Palo Alto Networks) benefit. Focus on emerging AI insurance and risk management companies. In a scenario where traditional retail and payment businesses are prolonged, existing large retail stocks become safe assets.

Key Triggers to Watch

Full Operation of Visa Intelligent Commerce: Transition from pilot to commercial deployment in Q2-Q3 2026. If successful, it will be an accelerating trigger for the agent economy.
First Large-Scale Agent Payment Incident: If erroneous orders or fraudulent use incidents by AI agents occur, it will lead to a pessimistic scenario of increased regulation and decreased trust.
n8n IPO / Major Acquisition: The IPO of a no-code AI automation platform will accelerate market recognition of the agent economy's infrastructure layer.
Full-Scale Deployment of OpenAI "Computer Use": If foundational models can fully autonomously perform web browsing and payments, the adoption of AI agents in the B2C sector will rapidly accelerate.
Gartner B2B Purchase Automation Interim Report: If the actual automation rate is confirmed in Gartner's report in early 2027, the market direction will be solidified.

Tracking Points

Next Trigger: Discovery of security vulnerabilities in AI agent payments and discussions on strengthening regulations in response (Q1-Q2 2027).

Continuation of this Pattern: Data privacy and ethical challenges in the AI agent economy: Transparency and accountability of algorithmic purchasing decisions.

Related patterns: Dragonfly Raises $650M — The Structure of VCs Contrarian in the Midst of "Extinction"

Sources: