The next major payments battle may not involve consumers swiping cards or scanning QR codes. Instead, it could be driven by artificial intelligence agents making purchases, paying for services and settling transactions without direct human intervention.
According to industry analysis, a potential $3 trillion to $5 trillion agentic commerce market by 2030 is beginning to take shape, prompting a race among established payment networks and crypto infrastructure providers to secure control over the next layer of digital payments.
At the centre of that contest are Visa, Mastercard and Coinbase, each pursuing different approaches to a future in which software, rather than people, increasingly initiates transactions.
The battle is about control, not just payments
While the debate is often framed as cards versus crypto, industry observers say the real contest is over who controls the mechanisms that sit behind transactions.
These include:
- Authorisation, which determines whether a payment is approved
- Tokenisation, which securely represents payment credentials
- Settlement, which moves money between parties
As AI-powered agents begin carrying out tasks such as booking travel, purchasing software services, paying for data access or handling routine business transactions, these systems become increasingly important.
The company that controls these functions could gain significant influence over how future commerce operates.
Mastercard wants AI agents to stay inside the card ecosystem
Mastercard has chosen a strategy focused on extending its existing payments infrastructure into the AI era.
Through its Agent Pay framework, the company aims to allow AI agents to transact using tokenised payment credentials while remaining subject to Mastercard's existing security, risk and authorisation systems.
The framework uses what Mastercard calls Agentic Tokens, which can be linked to specific spending rules.
Examples include:
- Merchant restrictions
- Spending limits
- Expiration dates
- Consent requirements
The approach enables AI agents to operate within an established payments ecosystem while preserving relationships with merchants, financial institutions and consumers.
Supporters view this as a practical path because it builds on infrastructure already trusted by businesses worldwide.
However, critics note that traditional card networks still carry transaction costs that may prove less suitable for extremely small or high-frequency AI-driven payments.
Coinbase is betting on a different future
Coinbase is pursuing a model built around crypto and stablecoin infrastructure rather than traditional card networks.
Its x402 payment framework focuses on enabling machine-to-machine transactions, particularly where payment values may be extremely small.
This becomes important if AI agents begin making frequent payments for:
- API access
- Computing resources
- Data services
- Digital content
- Software tools
In such scenarios, transaction costs become critical.
Industry observers note that if a payment is worth only a few cents, conventional card processing economics may become less efficient than blockchain-based alternatives.
At present, however, adoption remains limited. Reports suggest transaction volumes are still relatively low and include activity that may not yet reflect broad commercial demand.
As a result, the long-term viability of the model remains under observation.
Visa is building its own AI payment infrastructure
Visa is also moving aggressively into the emerging AI payments landscape.
Recent launches and infrastructure initiatives indicate that the company is developing tools designed to support automated and agent-driven commerce while maintaining the regulatory and security frameworks that have historically underpinned card payments.
Like Mastercard, Visa appears focused on adapting existing payment rails rather than replacing them.
The objective is to ensure that AI-powered transactions continue flowing through established financial networks even as buying behaviour evolves.
New signals suggest the race is accelerating
Several developments in recent months have reinforced the view that major players believe a new payments layer is emerging.
Among the notable moves:
- Stripe, Coinbase and MoonPay have introduced AI-focused payment infrastructure
- Mastercard has expanded its push into stablecoin infrastructure through acquisitions and partnerships
- Visa has launched new tools aimed at supporting next-generation commerce
- Open payment protocols are seeking to establish alternative transaction standards
Taken together, these developments suggest that companies are positioning themselves before widespread consumer adoption arrives.
For many industry participants, the goal is not simply to win market share today but to become the default infrastructure provider when AI-driven commerce reaches scale.
Why investors are paying attention
The significance of the battle extends beyond technology.
If AI agents eventually become responsible for millions or even billions of transactions, the payment networks supporting those transactions could become some of the most valuable pieces of digital infrastructure.
Investors are therefore watching several indicators closely:
- Growth in real-world AI transaction volumes
- Merchant adoption rates
- Expansion of tokenisation frameworks
- Stablecoin usage in commercial payments
- Production deployment of machine-to-machine payment systems
These factors may provide clearer evidence of whether agentic commerce is moving from experimentation to mainstream adoption.
The next phase of commerce may look very different
For decades, payments have been designed around people making purchasing decisions. The rise of AI agents introduces a fundamentally different model, where software could increasingly initiate, authorise and complete transactions independently.
That shift is creating new opportunities and new competitive pressures for payment providers.
Whether the future belongs to traditional card networks, stablecoin infrastructure providers or a combination of both remains uncertain. What is becoming clearer is that the contest has already begun.
As AI commerce infrastructure moves from concept to deployment, the companies that establish control over payment authorisation, tokenisation and settlement may ultimately shape how digital commerce functions for years to come.