There’s a quiet but profound transformation underway in the world of payments. It’s not a new checkout method or a sleek consumer-facing app—it’s happening at the infrastructure level, deep in the technical stack, where machine learning, automation, and natural language processing are beginning to rewrite the rules.

At the heart of this shift is Stripe and Adyen’s newly launched Model Context Protocol (MCP) tools designed to allow large language models (LLMs) to interact directly with their payment services. That might sound like another obscure technical upgrade—but it represents something much bigger: a potential turning point in how payments are built, modified, and accessed. One that could make the role of a payment engineer irrelevant.

What Is MCP and Why Does It Matter?

In simple terms, MCP is a way for AI systems—specifically LLMs like ChatGPT—to interact with payment API services on behalf of users. Instead of requiring someone to write and execute API calls manually (typically a developer or payments engineer), MCP lets a machine do the talking. The LLM acts as a translator between the human intent (a natural language prompt) and the technical execution (an API call), effectively automating the integration layer.

In Adyen’s recent proof-of-concept, this means a business user could ask for a PayByLink in a specific currency and amount through a platform like ChatGPT. The LLM interprets the request, asks any necessary follow-up questions, and then makes the correct call to Adyen’s MCP service to generate the link—all without the user needing to know a single line of code.

This isn’t the kind of “agentic commerce” often discussed in AI circles, where consumer-facing AIs search, compare, and purchase on our behalf. This is something different: enterprise-focused machine-to-machine collaboration, designed to streamline operations and empower non-technical users within businesses.

Why Have Adyen and Stripe Done This?

Adyen’s and Stripe move seems to be equal parts experimentation, customer enablement, and FOMO. Stripe debuted their capability at its Sessions event on 7th May, and Adyen followed on 10th June. Of the two players Stripe’s move is potentially more substantial because their customers have always been developer-centric. Adyen’s enterprise clientele are more likely to be interested in allowing business users to access services directly. Adyen is freeing up product and risk teams to take more direct control over their operations—without waiting in a dev queue.

A Changing Role for Payment Engineers

So, what does this mean for payment engineers—the professionals who’ve traditionally been responsible for crafting API calls, building integrations, and interpreting responses?

In short: their role is changing. Fast.

Traditionally, a payment engineer’s job involved translating business needs into technical reality. If a merchant wanted to enable 3D Secure for a subset of customers, the engineer would write a JSON request, initiate the API call, parse the response, and integrate the logic into a UX.

Now, with MCP, a business risk manager could simply ask: “Enable 3DS for high-risk transactions above £500.” The LLM would follow up with questions to clarify scope—Which cardholders? Which countries? What timeframe?—and then execute the required changes automatically.

No developer needed. No wait time. No sprint planning.

That’s both liberating and disruptive. On one hand, business users can operate more autonomously, tailoring services to their needs in real-time. On the other, engineers must reckon with a future where their traditional responsibilities are being automated or abstracted away.

Of course, we’re not talking about complete obsolescence. There will always be a need for experts to design and maintain the underlying systems, ensure security, and manage exceptions. But the nature of the work is shifting—from building integrations to orchestrating them, from coding to problem-solving, and from technical translation to strategic enablement.

Still Early Days—But the Direction Is Clear

It’s worth emphasising that we’re still early in this journey. Adyen’s POC is limited in scope, and while generating a PayByLink is useful, it’s a relatively simple task in the broader payments ecosystem.

That said, these things evolve quickly. Today it’s PayByLink. Tomorrow it could be onboarding a new payment method across markets, adjusting fraud thresholds, or setting up dynamic routing logic—all through natural language and AI-driven tools. And while merchants still need to consider the user experience (UX), even that is being chipped away by emerging LLM-based tools that can auto-generate front-end flows, brand elements, and logic trees in a matter of seconds.

We are entering the age of “vibe coding,” where a user’s intent—”make this checkout smoother,” “highlight this offer more clearly”—can be interpreted and executed by AI. As this progresses, the traditional lines between product, engineering, and business teams will blur, and a new kind of multi-disciplinary professional will emerge.

Not Death—But a Deadline

So, is the role of the payment engineer already irrelevant?

No. But it might be the end of the payment engineer as a purely technical executor. The days of being a gatekeeper to API-based functionality are numbered. The value is no longer in writing code, but in understanding the business context, customer behaviour, and commercial outcomes that the code is trying to achieve.

Engineers will need to evolve—fast. They’ll need to become more collaborative, more commercially literate, and more comfortable acting as strategists and systems thinkers rather than task-focused builders.

In that sense, MCP isn’t a threat—it’s a wake-up call. The future of payments will be built by those who can bridge AI fluency with commercial acuity, not just those who can write Python.

For the next generation of engineers, the message is clear: code is no longer the destination. It’s just one of many tools to get there.