Agentic commerce and programmable money are no longer distant concepts sitting somewhere on a five-year roadmap. In payments, the building blocks are already here – but the real question is whether banks, businesses, and regulators are ready to connect them in a way that is secure, scalable, and useful.

That was the core message from Mick Fennell, Business Line Director for Payments at Temenos, in a conversation with Mark Walker at Sibos in Miami. His view is clear: the technology is advancing, but the winners will be the institutions that modernize their infrastructure, manage risk well, and keep the customer experience simple.

In this post, you’ll learn what agentic payments actually are, what breaks first when banks try to implement them, how to think about the different types of digital money, and why the future of banking will likely be a mix of traditional and tokenized finance.

Agentic Payments Are Already Starting – But Mostly Behind the Scenes

The phrase agentic payments can sound futuristic, but the idea is straightforward: machines making decisions and moving value on behalf of people or businesses. Mick’s point was that AI agents are already being used across the payment process today. The real shift comes when those agents start interacting machine-to-machine, taking on more of the decision-making involved in what gets paid, when, and under what rules. For now, this is more advanced in business and corporate use cases than in consumer ones. That makes sense. Corporates often have more structured payment rules, more complex workflows, and a stronger need for automation across multiple systems. But there’s an important catch: agentic commerce does not work in isolation. It depends on the plumbing underneath it. That means banks need infrastructure that can handle:

• High speed and high volume
• Event-driven, API-first processing
• Cloud-native systems
• Clear controls around what an agent is allowed to do

If a bank is still heavily tied to legacy systems, the friction shows up quickly. The idea may be new, but the requirements are familiar: speed, reliability, and trust.

What Breaks First: Infrastructure, Risk, and Customer Controls

When banks start thinking about agent-to-agent or computer-to-computer transactions, the first failure point is usually not the idea itself. It’s the infrastructure. Mick broke this down into two major layers. First is the technology stack: can it handle the required speed, performance, and capacity? Second is the control layer: can the bank manage the risk that comes with automated decision-making? That second layer matters just as much as the first. If an AI agent is making payments, the bank still has to know:

• What the agent is allowed to buy
• Which product types it can use
• What volume or threshold should trigger controls
• How to monitor behavior in real time
• What service level the customer has agreed to

In other words, the bank still needs to act like the trusted gatekeeper. The difference is that the customer may no longer be clicking every button manually. Mick compared this to direct debits and collections. Banks already manage rules, mandates, and limits in those systems.

Agentic payments are not entirely new territory – they are those same concepts, but operating at a much higher speed and volume. That’s why this moment is less about inventing a new banking model and more about upgrading the current one. A helpful way to think about it is this: the bank is not replacing trust with automation. It is trying to automate only the parts of the process that can be safely governed.

The Four Types of Digital Money Banks Need to Understand

One of the most useful parts of the conversation was Mick’s simple framework for digital money. He grouped it into four categories and used the Four Musketeers as shorthand:

1. CBDCs – public money

Central bank digital currencies are issued by the regulator or central bank. Mick described them as the most stable and straightforward option – the “moral compass” of the group. For banks, CBDCs are important because they sit close to public trust and sovereign control. But at the same time, they are still relatively limited in real-world deployment. There have been plenty of pilots and proofs of concept, yet adoption has been slower than some expected.

2. Stablecoins – private money

Stablecoins are issued by private players and are the most visible form of tokenized money in many current discussions. They are also where much of the energy in the market is right now. Mick framed them as bold, fast-moving, and useful – but with caution attached. They can support interesting use cases, particularly for large organizations and multinational flows, but they also come with risks around liquidity, backing, and regulation.

3. Tokenized deposits – bank money

This is where the practical value for banks becomes especially clear. Tokenized deposits are customer deposits represented in tokenized form so they can move on blockchain-based rails while still sitting inside the bank’s regulated environment. Because this is still bank money, it aligns naturally with existing rules and relationships. For Mick, this is one of the strongest near-term use cases because banks already manage this money. Tokenization simply gives them a new way to use it.

4. Cryptocurrency – higher risk, less certainty

Crypto sits in a different category. Mick described it as the most speculative of the group – exciting, but significantly more uncertain in terms of backing and risk. That doesn’t mean there is no use case. It means banks need to be much more selective and realistic about where it fits. The bigger point is that banks can’t treat all digital assets the same. They need a strategy for each type, based on:

• Legal and regulatory risk
• Customer demand
• Redemption under stress
• Operational impact
• How much trust the bank is willing to extend

Why Cross-Border Payments Still Depend on the Same Old Reality

A lot of the excitement around stablecoins and tokenized money comes from cross-border payments. But Mick was careful to ground the conversation in how money actually works today. His point was simple: money doesn’t really move borders – currencies change hands between systems .That’s why cross-border payments have always involved rails, liquidity pools, correspondent relationships, and conversion into local currency on the other side. Stablecoins can simplify parts of that process, but they don’t magically remove the need for liquidity or local settlement. This is where the strongest use cases often appear first:

• Large corporates moving value between entities in different countries
• Multinationals managing treasury more efficiently
• Banks operating inside their own networks across regions
• Internal transfers where the institution can control both ends of the flow

If a bank in one market is moving value to itself in another market, tokenized deposits may make the most sense. If it’s interacting across different institutions or jurisdictions, stablecoins may be useful – but only if the surrounding infrastructure supports them. That’s the real bottleneck. Not the coin itself. The rail-on and rail-off infrastructure matters just as much as the digital asset in the middle.

And that leads to a crucial operational question: how many liquidity pools can a bank realistically support? Mick pointed out that if the number of stablecoins keeps growing, banks risk fragmenting liquidity across too many instruments. That creates balance-sheet pressure and funding complexity. It’s not only a technology issue – it becomes a treasury and risk-management issue very quickly. The Future Bank Will Run Traditional and Tokenized Finance Side by Side

If there was one big strategic takeaway from the conversation, it was this: the future of banking is likely to be hybrid .Banks will not suddenly abandon fiat systems, existing payment rails, or conventional deposits. Instead, they will need platforms that can support both traditional and tokenized finance in one place. That matters because customers do not want complexity. They want outcomes. As Mick noted, the user experience still comes back to four things, which mirror the G20’s priorities for better cross-border payments:

• Cost
• Access
• Speed
• Transparency

These are the same factors that matter whether you’re a retail customer, a corporate treasury team, or a bank moving funds on behalf of a client. Tokenized money, smart contracts, blockchain, and distributed ledger technology can improve speed and transparency right away. Access also improves because programmable money can support new workflows and new types of automation. Cost is trickier. The underlying asset may be cheap to move, but the ecosystem around it – liquidity, conversion, rails, compliance, and operations – still adds cost. That will come down over time, but not overnight. Mick also made an important point about scope: tokenization is not just about payments. It can also extend into:

• Lending
• Trade finance
• Securities and digital assets
• Holdings and other banking products

So the challenge for banks is not just “Should we support tokenized money?” It is “How do we build a single platform that can manage every form of value as the market evolves?” That is exactly where strategic infrastructure providers see the opportunity.

What Banks Should Do Next

The hardest part of this transition is not understanding the buzzwords. It is deciding what to do now. Mick’s view is that banks should think in terms of capability windows : what do they need in 6, 9, 15, 18, or 36 months? The answers will vary depending on market maturity, regulation, and customer demand. Here’s the practical version:

1. Assess your infrastructure

   • Can it handle real-time processing?
   • Is it cloud-native, API-first, and event-driven?
   • Or will legacy systems create bottlenecks?

2. Define your risk controls

   • What can an agent do?
   • What thresholds need monitoring?
   • What limits should apply by customer, product, or amount?

3. Choose your digital money strategy

   • Decide what role, if any, CBDCs, stablecoins, tokenized deposits, and crypto should play.
   • Do not assume all four deserve equal attention.

4. Design for interoperability

   • Customers do not want to know what rail is used.
   • They want money to move safely, quickly, and transparently.

5. Build for regulation, not around it

   • Different markets are moving at different speeds.
   • The regulatory rails are not uniform, so a one-size-fits-all approach will fail.

The banks that move early will not necessarily be the ones chasing the loudest trend. They’ll be the ones putting the right building blocks in place before the market forces their hand.

---

Frequently Asked Questions

What are agentic payments?

Agentic payments are payments initiated or managed by AI agents that can make decisions and act on behalf of a person or business. They are most powerful when paired with rules, controls, and infrastructure that allow machine-to-machine transactions safely.

Are programmable money and tokenization already in use?

Yes, in some use cases. Banks and corporates are already testing and deploying parts of these models, especially in tokenized deposits, stablecoin experiments, and broader digital asset infrastructure. The market is still early, but it is no longer theoretical.

Why are tokenized deposits important for banks?

Tokenized deposits let banks use customer deposits in a tokenized form while keeping them within the existing regulatory and institutional framework. That makes them a strong near-term option for improving efficiency without abandoning trust or control.

Will stablecoins replace traditional money?

Probably not. Stablecoins may become useful in certain corporate, cross-border, and internal treasury scenarios, but they still depend on liquidity, regulation, and supporting rails. In most cases, they are more likely to coexist with traditional money than replace it.

What should banks prioritize first?

Banks should start with infrastructure readiness, risk controls, and a clear strategy for which digital money models they want to support. The technology matters, but the bigger differentiator is how well a bank can protect trust while improving speed and usability.

---

The future of payments is not just about faster money – it is about smarter infrastructure, better controls, and platforms that can support both traditional and tokenized finance. Banks that prepare now will be better placed to manage risk, serve customers, and capture the upside as programmable money becomes more practical. If you want to explore the broader transformation in banking, the next step is to look at where your current systems will hold you back and where they can evolve.4. One-Sentence Summary Banks should prepare for a hybrid future where agentic payments, tokenized deposits, stablecoins, and traditional rails coexist under stronger infrastructure, risk controls, and customer-focused design.

 

The post News & Views Podcast: The Real Barrier to Agentic Payments Is Not the AI appeared first on The Fintech Times.