Innovation was everywhere. But what struck me most was that the conversation has moved well beyond simply asking what is possible?

Across payments, digital assets, artificial intelligence and the infrastructure supporting international commerce, the questions are increasingly about how we turn innovation into something scalable, resilient, trusted and commercially sustainable.

Five themes particularly stood out for me:

The evolution of cross-border payments continues at pace. Competition for international payment flows is intensifying, with banks, fintechs and other providers investing significantly in faster, more transparent and increasingly digital propositions. One clear theme at Sibos was the growing role that domestic instant-payment capabilities can play within international payment journeys, including as the final leg of a cross-border transaction. But speed is only one part of the equation.

The characteristics that have always mattered to customers have not disappeared as payments have become faster. The winners won’t simply be the fastest. They’ll combine speed with transparency, certainty, resilience and trust.

The conversation around tokenised deposits and digital assets felt noticeably different this year.

There is still substantial experimentation, but this is no longer a discussion based solely on concepts and pilots. Across the industry there are tangible initiatives, propositions and solutions operating today, giving us a much clearer view of how tokenised forms of money could become part of the future financial system. At the same time, different markets are taking different routes. Commercial innovation, regulatory approaches, public-private collaboration and interoperability are developing in different ways and at different speeds.  

That means the next stage is perhaps less about proving that the technology works and more about proving that it can scale commercially and operationally. What are the sustainable business models? How do solutions interoperate? How do we create sufficient network effects? And, perhaps most importantly, the question is shifting from whether tokenisation works to whether it solves a genuine customer problem better than today’s infrastructure. Those questions will increasingly determine which initiatives move from innovation to meaningful adoption.

One of my strongest takeaways from Sibos was just how significant the infrastructure challenge facing the industry has become.  Banks and payment providers are operating in an environment where customer expectations continue to increase, technology continues to advance and the requirement for resilient, secure and always-on services becomes ever greater. That creates a fascinating tension.

There is a continued need to invest in existing infrastructure, making today's payment systems faster, more resilient, more connected and easier for customers to use. At the same time, new infrastructure is emerging, particularly around digital assets and tokenised money, creating choices about where and how organisations should invest for the future. Sometimes these paths converge. At other times they widen.  ISO 20022, APIs, instant payments and developments in digital money are not simply individual technology programmes. Collectively, they are reshaping the infrastructure choices available to the industry. Far from infrastructure becoming less of a differentiator, I think the opposite could prove true. The advantage may lie in making the right investment choices: modernising today’s capabilities while preparing for tomorrow’s and connecting those worlds effectively could become an increasingly important source of differentiation. 

AI was inevitably one of the dominant conversations at Sibos, but the discussion is already evolving. AI is moving beyond productivity. The bigger question is what happens when machines begin to participate directly in commercial and financial decision-making?  Agentic commerce presents the prospect of AI agents performing activities that have traditionally required human intervention, from procurement and treasury processes through to initiating or influencing payments.  If that develops at scale, the consequences could extend far beyond operational efficiency.

Increasing automation and the disintermediation of humans from some payment decisions could change the speed and patterns of commercial activity. It could influence liquidity requirements, cash-flow forecasting and balance-sheet management, and potentially alter some of the dynamics within financial markets themselves. That creates considerable opportunity, but equally significant questions. Guardrails, identity, authority, transparency, accountability and interoperability will all matter. But more broadly, this feels like an industry question rather than something any one institution can address independently.  

The international transaction banking ecosystem is becoming increasingly complex.  There have never been more choices in how money can move, how transactions can be financed, how liquidity can be managed or how clients can connect into the financial system. And each new capability can create another potential route through the ecosystem. Interestingly, the fundamental outcomes customers care about have changed relatively little. Speed, transparency, resilience, certainty and cost have stood the test of time.  What is changing is the number of ways in which those outcomes can be delivered.  Different infrastructures and propositions can optimise for different priorities. One solution may offer greater speed, another broader reach, another reduced cost or increased certainty. The result is not necessarily one industry-wide destination, but potentially multiple paths depending upon the customer, use case and priority.

For us at NatWest, that reinforces something fundamental about transaction banking.  We first need to truly understand what the customer is trying to achieve, and which outcomes matter most to them. We then need the breadth of capabilities and services to translate those priorities into the right solution. In a world of increasing choice, the differentiator may not be having more options. It will be knowing which options matter, for whom, and when. 

Sibos 2026 was a fantastic opportunity to bring many of these themes to life. What makes the period ahead so exciting is that these developments don't exist independently. Changes in payment infrastructure influence the potential for tokenised money. AI could change how and when transactions are initiated. New technology creates new choices, while simultaneously increasing expectations around resilience, transparency and trust.

Our challenge as an industry is to connect those dots.

At NatWest, we're continuing to work through these themes and what they mean for our customers, our capabilities and the future of transaction banking. We'll be sharing more content over the coming months, taking a deeper look at a number of the areas discussed here. And the conversation certainly won't stop with Sibos. We're also starting to shape our NatWest Future of Payments conference for 2027, where I expect many of these questions will continue to resonate.

There is no shortage of innovation ahead. The real opportunity, is turning possibility into progress: building solutions that customers trust, that solve genuine problems and that can operate at scale. That is where the next chapter gets really interesting.