Key Takeaways
- Agentic commerce inverts the very assumptions of payment architecture. Sessions persist for days, and authorization spans fragmented networks, so a translation layer that absorbs these differences becomes necessary.
- Adyen's single platform, which controls acquiring, processing, and settlement end to end, is structurally suited to this translation role. That is precisely why it can claim the title of universal translator.
- Whereas Visa and the networks operate at the authentication and token layer, Adyen aims for an intermediary layer that sits between merchant systems and agent platforms.
A Payment Processor Calling Itself a Translator
Adyen, a major payments platform, describes its role in agentic commerce as that of a universal translator. Fintech analyst Sam Boboev uses this metaphor as a starting point to read a structural shift in payment infrastructure. This article follows his analysis to ask why a payment processor would aim to become a translation layer.
The evolution of digital commerce has arrived at a critical structural pivot. By Sam Boboev, Fintech Wrap Up.
www.finextra.comBoboev's argument is clear. Digital commerce has lived through several revolutions of the screen, including mobile, social commerce, and marketplaces. Each one repainted the front end that customers touch, yet the transactional architecture underneath was left largely intact.
Agentic commerce reverses that premise. Software agents act autonomously on behalf of users, and sessions stay stateful across days or even weeks. Identity, trust, and payment authorization must now be managed over fragmented, asynchronous networks. This is not a surface change but a rebuild of the plumbing itself.
The product launch of Adyen Agentic, announced on June 16, 2026, is covered in a separate article on this blog. Here we deliberately avoid repeating the product specifications and focus on the structure: why a translation layer is needed, and why Adyen is suited to provide it.
The Bottleneck Is Transaction Infrastructure, Not the LLM
Many people assume agentic commerce has not taken off because AI is not yet smart enough. Boboev rejects this directly. The rate-limiting step is not the capability of large language models but the absence of scalable transaction infrastructure.
The bottlenecks inside enterprise systems run deep. Protocol fragmentation, product data structures unusable by agents, rigid legacy checkout stacks, the trust paradox of distinguishing legitimate agents from malicious automation, and the engineering overhead of bespoke per-platform onboarding. None of these disappears by making the model smarter.
The scale of the market explains why the problem cannot be ignored. McKinsey and Bain project that the global agentic commerce market will reach between 3 trillion and 5 trillion dollars by 2030. Accenture estimates that over 30% of online commerce will run through AI agents by 2030, representing close to 3.1 trillion dollars in transactions.
What is decisive here is how differently data behaves for an agent. A human buyer can wait, refresh, or ask a question when a description is ambiguous. For an autonomous agent, data ambiguity means a hard programmatic failure. A mere ten-minute lag in inventory can halt the execution of a transaction.
Why the Logic Leads to a Translation Layer
As AI platforms launch new purchasing surfaces one after another, merchants face an awkward choice. Each platform uses a different protocol, demands a different product data format, and follows different rules for cart creation and checkout. Every time a new surface appears, the merchant has to launch an integration project from scratch.
Left unchecked, this structure leads to a proliferation of closed gardens. Each platform tightens its enclosure, and merchants are forced to either bet on one or burn resources serving them all. Neither outcome is healthy.
The idea of a translation layer is born to remove the structure of the bet itself. The merchant integrates once, and a layer in the middle converts that single integration into every agent platform, protocol, and payment method. Like an interpreter mediating meaning between speakers of different languages, it translates transaction parameters between merchant systems and the world of agents. Adyen positions this layer as an open abstraction layer that does not favor any particular ecosystem.
What matters is that the translation layer is more than an API converter. At discovery it shapes product data into a machine-readable form, at the cart stage it handles constantly shifting session state, and at payment it protects merchant sovereignty. It absorbs differences of distinct character at each stage, consistently.
Why a Single Platform Suits a Translator
So why, among many payment companies, can Adyen claim this role. The key lies in the single platform strategy the company has long championed.
In the conventional payments world, gateway, risk management, processing, acquiring, and settlement are each handled by a different vendor. Merchants stitch together several specialists just to build one payment flow. Adyen turned this convention on its head and folded these layers into a single product experience. By connecting directly to card networks and obtaining its own banking and acquiring licenses in key regions, it controls everything from acquiring to processing to settlement end to end.
This end-to-end structure is decisive for a translation layer. When an agent moves across diverse purchasing surfaces, multiple payment methods, and regional regulations, absorbing those differences requires reach into every stage of the payment lifecycle. A provider handling only one stage cannot translate fragmentation from end to end. Seeing the whole transaction through a single platform becomes the precondition for being a translator.
Adyen's Agentic Payments is built on top of the existing platform that has processed trillions of dollars annually. Its tokenization, authentication, and fraud systems can be reused as is, so there is no need to lay fresh plumbing for agents. The accumulation of the single platform converts directly into the reliability of the translation layer.
The Battle Over Merchant Sovereignty
The heaviest point in any discussion of the translation layer is the risk of merchant disintermediation lurking at the payment execution stage. Boboev frames this as a severe strategic threat.
Suppose the intervening AI platform takes hold of the payment vault and handles checkout directly. The merchant is demoted to a commoditized fulfillment function, forced to compete only on speed and price. The customer relationship, brand continuity, and sovereignty over transaction data could all be drawn up into the platform.
What Adyen advances at this stage is the idea of an explicit and verifiable mandate. An agent cannot execute a transaction unless it presents a cryptographically signed authorization proving the user's intent, spend limits, and temporal boundaries. This contract protects the merchant from buyer's remorse disputes and serves as a key input to risk models.
Adyen also lets merchants manage their own tokenization lifecycle, avoiding platform-specific lock-in. In place of cookies and logins that break down in conversational interfaces, it uses a stable universal payment token that works across fragmented surfaces. It then keeps settlement, compliance, and the post-purchase customer relationship within the merchant's domain, preserving merchant-of-record status. The stance is to translate without handing over control.
Where It Sits Among Payment Players
Agentic payments have been a fierce battleground since players moved in unison in 2025. Yet the layer each company targets is not the same. Adyen's distinctiveness stands out precisely in this comparison.
Visa and Mastercard approach from the network layer, verifying agent identity and underwriting signed intent on the network. Stripe builds payment primitives that let agents use a payment method without exposing credentials. OpenAI and Google try to define the conversational protocol itself. Protocols like AP2 list more than 60 organizations, including Adyen itself.
| Player | Layer Targeted | Approach Focus |
|---|---|---|
| Adyen | Intermediary layer between merchants and agent platforms | Translates one integration into every platform, protocol, and payment method |
| Visa / Mastercard | Card network layer | Verifies agent identity and underwrites signed intent on the network |
| Stripe | Payment primitive layer | Lets agents use a payment method without exposing credentials |
| OpenAI / Google | Conversational protocol layer | Standardizes the transaction language from discovery to checkout |
Laid out this way, Adyen's position becomes clear. Neither network nor protocol, it takes on the intermediary layer that sits between merchant systems and agent platforms. Where others handle part of a layer, Adyen bets on translating the merchant's single integration in every direction. Rather than trying to guess the winner of the protocol war, it takes a position where the merchant is fine no matter the outcome.
Implications for Merchants and Payment Teams
This structural shift carries concrete implications for on-the-ground preparation. First, agent-led payments are no longer a someday concern but have entered a stage that demands integration decisions now. Given the market scale projected by McKinsey and Accenture, the opportunity cost of waiting is not small.
Getting product data in order is the first step. Since data ambiguity means a halted transaction for an agent, holding a machine-readable, fresh catalog becomes the entry ticket at the discovery stage. This connects directly to AEO (AI Engine Optimization) work.
The criteria for choosing a processor also change. The question is not only whether fees are low. Can it cover diverse platforms and protocols with a single integration. Does it leave mandate and token sovereignty with the merchant. Does it avoid disintermediation and preserve merchant-of-record status. Choosing a translation layer also means choosing these conditions of sovereignty.
Conclusion
When Adyen calls itself a universal translator, it is a declaration of structure rather than rhetoric. As agentic commerce inverts the premises of transaction architecture, real value emerges for a translation layer that absorbs fragmented differences. A single platform that controls everything from acquiring to settlement is structurally suited to that role.
As payment players divide the layers among themselves, Adyen has moved to claim the space between merchants and agents. What merchants and payment teams must do is move up their preparation for agent-led payments and revisit data readiness and processor selection through the lens of preserving sovereignty. The competition over the translation layer is also an occasion for merchants to choose their own position anew.
