Issuu

28 September ˜ 1 October 2026 | Miami, Florida USA

Let’s get real BNP Paribas shares its learnings from live DLT

Double agent Harness the power, contain the risk: Red Hat’s AI playbook

Integrate to accelerate How interoperability will drive success for Bottomline

The

Future Finance of

THE DAWN OF AI-DRIVEN ECONOMIES: WAKING UP TO THE TOKENISED SUNRISE

Digital finance for AI-driven economies

28 September – 1 October Register now on Sibos.com

DISCOV

S6 O SM IIAB 2 0 2 I M

DISCOVERSIBOS 6 22 INFRASTRUCTURE

DLT

Rules Britannia! The great payments renewal

The future is here (almost)

EUROPEAN PAYMENTS

Driving open banking forward

The Italian paytech CBI has a key role in taking A2A payments mainstream – not just in Italy

ARTIFICIAL INTELLIGENCE

The new AI playbook

Open source enterprise solutions provider Red Hat addresses the Top 3 challenges

SECURITIES

Converging on co-existence

CSD Euroclear maps out its approach to a twin-track world of TradFi and tokens

CROSS-BORDER

The ‘new’ correspondent banking system Tokenisation and DLT are not incompatible with using the established network of intermediary banks in wholesale cross-border settlement. Deutsche Bank has a foot in both camps

PAYMENTS

Navigating change safely

The world can feel like a very small and unpredictable place for businesses engaged in global trade. So how helpful is J.P. Morgan Payments’ tech in steering them through troubled waters?

BNP Paribas’ Securities Services business has moved beyond the sandbox and kicked the tyres of tokenisation in the real world. What has it taught the team so far?

PAYMENT RAILS

Upgrade the tech, but trust still moves the money Innovation will be everywhere at Sibos in Miami. But for Simon Eacott of NatWest, real progress happens where old meets new

POST-TRADE

Reconciled to the truth

As settlement cycles shrink and financial institutions wrestle legacy infrastructure, Smartstream believes only one type of intelligence can handle the pressure

NEW ARCHITECTURES

No more ‘business as usual’

After decades of disruption, the payments industry has accepted that change is the new normal. But dynamic interoperability can help absorb the shifts, says Bottomline

TREASURY

Setting liquidity free

In helping to release treasury teams from the drudgery of manual processes, J.P. Morgan Payments is also liberating the cash that can be crucial to growth

THE EDITOR’S VIEW

Staring out into the vast Atlantic Ocean from Miami Beach, it’s impossible to see the horizon clearly. That’s kind of the problem Sibos delegates have struggled with when it comes to distributed ledger technology. We’ve all been squinting at a hazy DLT skyline for years. There was the occasional crypto baron’s superyacht sailing across a tokenised sea, but even though the rest of the financial industry seemed to have been rowing towards them for decades, real-world applications of the tech beyond token trading never seemed to get any closer. Last year’s conference in Frankfurt boldly put the technology front and centre, though, and in our interviews for this supplement it’s clear that momentum hasn’t slowed. The industry is impatient for progress. But it’s also getting real. There is less talk of wholesale change, radical realignment. The direction of travel now is towards peaceful co-existence and convergence between the existing banking system and the new distributed architecture. The most obvious demonstration of that is the European-led Project Agorá – which this year proved atomic settlement can happen safely and without the traditional players relinquishing their autonomy. With so many of them now on the same page, let’s hope we cross the horizon soon. Sue Scott, Editor

DISCOVER SIBOS MAGAZINE EXECUTIVE EDITOR Ali Paterson

ART DIRECTOR Chris Swales

GENERAL MANAGERS Tom Dickinson Chloe Butler

SUB EDITOR Frank Tennyson

CHIEF REVENUE OFFICER Shaun Routledge EDITOR Sue Scott

PHOTOGRAPHER Jordan Drew PARTNERSHIP MANAGER Georgia Stubbs

4 DISCOVERSIBOS2026

PRODUCTION TEAM Tom Dickinson Matthew Burls Dylan Young ONLINE EDITOR Lauren Towner VIDEO TEAM Louis Jean La Grange David Hill Ethan Peck

FEATURE WRITERS Hannah Duncan David Firth Tracy Fletcher James Grant Natalie Marchant Sue Scott James Tall Frank Tennyson

MIAMI 2026 The Fintech Magazine is published by FF NEWS LTD. Pantiles Chambers 85 High Street Tunbridge Wells, TN1 1XP CONTACT US ffnews.com

DESIGN & PRODUCTION www.yorkshire creativemedia.co.uk IMAGES BY www.istock.com PRINTED BY Everything Creative "PROUDLY NOT ABC AUDITED"

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, photocopying or otherwise, without prior permission of the publisher and copyright owner. While every effort has been made to ensure the accuracy of the information in this publication, the publisher accepts no responsibility for errors or omissions. The products and services advertised are those of individual authors and are not necessarily endorsed by or connected with the publisher. The opinions expressed in the articles within this publication are those of individual authors and not necessarily those of the publisher.

FFNEWS.COM

DISCOV

S6 O SM IIAB 2 0 2 I M

INFRASTRUCTURE

As the key utility operator, Pay.UK must continue to keep systems running safely and efficiently while the country redesigns its payments infrastructure. David Crawford considers some of the difficult questions the future poses Payments infrastructure is at its most successful when nobody notices it. Salaries arrive, bills are collected and money moves between accounts in seconds, with the complex machinery underpinning those transactions largely invisible to the people and businesses relying on it. At the heart of that machinery in the United Kingdom sits Pay.UK. As the operator of the nation’s core retail interbank payment systems – including Faster Payments, Bacs and the Image Clearing System – the organisation occupies a critical position in one of the world’s most sophisticated payments markets.

6 DISCOVERSIBOS2026

But its remit stretches beyond keeping the rails running. Pay.UK also operates services designed to help prevent fraud, including Confirmation of Payee, which now covers more than 99 per cent of organisations initiating Faster Payments in the UK and helps people check they are sending money to the intended recipient. According to recent UK Finance figures, UK losses from authorised push payment fraud reached £576.4million in 2025, up 19 per cent year-on-year, illustrating the reality that faster, more seamless payments must be accompanied by advances in security. That’s top of mind as Britain embarks on a major overhaul of its payments landscape. The government’s National Payments Vision and 2026 Payments Forward Plan have placed innovation, competition and security at the centre of the agenda, while the Retail Payments Infrastructure Board is helping shape the next generation of UK retail payments. For Pay.UK, that creates a balancing act: maintaining the resilience and security of systems used across the economy today while preparing for what comes next. We spoke to Pay.UK’s Chief Strategy and Transformation Officer, David Crawford, about modernising Britain’s payments infrastructure, tackling fraud, fostering innovation and what the next chapter of UK payments could look like.

David Crawford, Chief Strategy & Transformation Officer at Pay.UK

DISCOVER SIBOS MAGAZINE AI is rapidly changing financial services. What role could it play in helping Pay.UK make the payments system more resilient and identify financial crime before it happens? DAVID CRAWFORD First and foremost, given our priority to keep payments flowing no matter what, I think about resilience. It’s certainly something we

Sterling effort: Pay.UK helps build resilience, security and interoperability into the system

FFNEWS.COM

take huge pride in – maintaining that resilient environment not just within Pay.UK, but across the entire ecosystem. So, it’s about using AI to help us be even more resilient than we already are; by helping us scan the environment every minute of every day for signals and threats. Through analysing those signals, we could detect and prevent incidents before they even happen. And if we can’t prevent something, AI can help us scenario-plan virtually instantly, so that we already have the playbooks and mechanisms to respond the next time. Then there’s the data within payments themselves. How and should we start including information about whether a payment originated from AI, an agent or a human? And what would that mean for how we authenticate or authorise that payment? DSM Your previous answer touches on agentic commerce, which could fundamentally change not only who initiates payments, but also how frequently transactions take place. Are today’s payment rails and authentication systems ready for AI agents? DC You mean, do we have the digital motorways to carry the traffic we expect to see over the next three to five years from agentic commerce? I think the answer is broadly yes. Although we’re not seeing that huge increase in volume right now, we have to be ready for it. Consumers probably won’t spend much more money, but what they do spend could become increasingly fragmented. Instead of doing a weekly food shop, for example, your fridge might start ordering things itself. We’re going to see more payments as a result. Your other question is really about whether we need to change the rules of the road. If a payment was made by an agent, what does that mean for authorisation, and what does it mean for liability? If an agent sends your life savings to a scammer, who’s liable? Is it you because you authorised the agent to act for you, or is it the AI company hosting that agent? These are really difficult questions that we’re going to have to face. At the same time, we’re excited about capabilities such as programmable payments, programmable money and the linkage between new forms of money and new payments infrastructure. I can’t tell you what all those user journeys will look like, because I don’t think they’ve been invented yet! DSM There are decades of accumulated rules,

standards and obligations underpinning

FFNEWS.COM

payment schemes. How can Pay.UK strip away that complexity and make payments genuinely interoperable? DC Just about every scheme in the world has evolved over time and ended up with rulebooks that can contain thousands of obligations and rules. So, when we think about cross-border, cross-scheme and cross-forms of money, enabling all of those to operate seamlessly is going to be a huge challenge. We’ve spent a lot of intellectual capital and energy looking at how we solve that problem. We’ve taken our own rulebooks, extracted the rules, deduplicated and simplified them, and we want to take that to the nth degree. We don’t know what good looks like yet. Is it half the number of rules we have today? Is it a tenth? But we want to get down to the bare minimum because that makes it much easier to establish the conditions for seamless transfers of value between countries, schemes and forms of money. We want to go further and move away from recording these things in documents, too. We want to digitise the rulebook so, as an operator, you could ask ‘Which rules matter for me when onboarding?’, ‘Which matter for cross-border payments?’, ‘If I take a certain action, does that satisfy those rules?’. The complexity exists because, as you say, these systems have evolved over many years. Bacs, for example, is almost 60 years old. As we’ve added things to keep payments safe, resilient and efficient, we’ve also built in complexity. We need to simplify that for everyone.

There’s a dual threat and opportunity that AI affords everyone – us, our participants and consumers. We’re already seeing AI being used across the ecosystem to protect against fraud, whether that’s by banks, ourselves or consumers using apps that can help keep them safe. But we have to be alive to the fact that those same tools can be used against us. That means thinking about how we build our defences to manage that threat. Pay.UK cannot simply participate in that discussion. There are times when we really have to lead it – bring people together and create the conditions that allow the industry to solve these problems collectively. DSM The UK now has an opportunity to build its next generation of payments infrastructure around needs that barely existed when today’s systems were created. What will modernisation actually change for consumers, merchants and businesses? DC Building resilience by design is really important. But how we build fraud prevention in by design is going to be critical. Modern infrastructure can take user journeys we’ve already started developing – things such as account-to-account payments – and make them seamless, digital and straightforward, including bringing account-to-account payments to the physical checkout. It will also offer users a 24/7 capability. Being able to undertake payments without cut-offs being imposed at various times of day will be really important. Then there are new capabilities such as programmable payments and money. I think those will unlock innovation for various parties – not just ourselves and banks, but also third and fourth parties that can create new experiences for end users. That’s what I’m personally most excited about: seeing people grab those capabilities and develop ideas and innovations that we haven’t even thought about today. If I go back 10 or 15 years, I don’t think anybody envisaged QR codes, splitting a bill digitally or sending somebody a request to pay you £10 for something you bought yesterday. People innovated on the existing infrastructure to make those things possible. I think the next generation of innovation is going to be really exciting – we just don’t know yet exactly what it will bring.

As we’ve added things to keep payments safe, resilient and efficient, we’ve also built in complexity. We need to simplify that for everyone

DSM Real-time, AI-driven payments also mean threats can emerge and spread at unprecedented speed. Does protecting the next generation of payments require the industry to become much more collaborative with its data? DC Yes. Moving towards real-time environments fuelled by AI means we can no longer operate in our individual silos. We need to collaborate, share information and share our understanding of threats – right down to sharing data with each other. We are already exploring the Enhanced Data Exchange with our Pay.UK participants. It’s a little like Confirmation of Payee, which we run, and it addresses how we can share additional information that helps us detect and prevent fraud before it happens.

DISCOVERSIBOS2026 7

DISCOV

S6 O SM IIAB 2 0 2 I M EUROPEAN PAYMENTS

The Italian paytech CBI has a key role in taking A2A transfers mainstream – not just in Italy but throughout Europe, and beyond

CBI’s latest product to accelerate that strategy is Request to Pay, launched in June 2026 under the SEPA Request to Pay (RTP) messaging scheme for message transfers. Ian Horne spoke to CBI’s Head of Marketing and Sales, Alessio Castelli, about Europe’s payments vision, how CBI tools can boost customer satisfaction at home and enhance cross-border interoperability.

CBI is a critical part of Italy’s financial landscape developing innovative services for corporate and retail end clients.

DISCOVER SIBOS MAGAZINE How has Europe’s Instant Payment Regulation (IPR) and mandatory Verification of Payee (VoP) reshaped Italian payments? ALESSIO CASTELLI The IPR went live last October in Europe. It required payment service providers (PSPs) to offer instant credit transfers at the same price as traditional credit transfers with the aim of increasing account-to-account payments. A PSP must also enable a payer to verify if the IBAN to which they are sending a payment is correctly associated with the name of the payee. Within this landscape, CBI launched Name Check, which is a VoP service that has already been adopted by more than 250 banks in Europe.

A payment standards setter, open finance ecosystem aggregator and a centralised interbank processor, CBI’s impact on the Italian payment system has been profound. And its influence isn’t confined to its home market. The bank-owned utility provider’s shared, interoperable infrastructure is increasingly seen as key to helping Europe strengthen its financial sovereignty by making digital payments fast, safe and cheap to use. That, European leaders hope, will increase the adoption of digital payments and account-to-account solutions. 8 DISCOVERSIBOS2026

Alessio Castelli, Head of Marketing and Sales at CBI

We register about 200 million verifications each month, and almost 70 per cent of VoPs turn up a match result. Among the no-match or close-match results, there will be some fraud attempts or misdirection of payments, so we are continually enhancing our solutions. This includes value-added services, such as the so-called Trade Name, which collects all the commercial names of the payee in order to reduce ‘no matches’ output. Bulk VoP is another solution that relieves the effort related to the unboxing and verification of single lines of payments included within a bulk file. It’s also worth highlighting IBAN Lookout. That is a service that is able to detect IBAN codes that have been flagged as suspicious by the financial community. So, you may have a correct match between an IBAN code and the name of the payee, but IBAN Lookout will tell you if this payee has also been highlighted as being suspicious. CBI had already centralised many functionalities to allow efficiency and cost reduction to all the banks in its financial community. The VoP requirement changed the behaviours and habits of users, so FFNEWS.COM

customers – to receive payment requests via internet banking. Because it is a simple message between a biller and a final user, it is very useful for all the actors in the value chain. PSPs can maintain the last-mile relationship with the final users, which also gives them the opportunity to cross-sell and upsell products, while corporates can offer customers a new payment instrument, which has been very well received. In countries where Request to Pay has been heavily adopted, we’ve seen a customer satisfaction increase of 85 per cent, in relation to traditional – and often limited – payment instruments. For payment service users, this new payment solution offers unique opportunities. For instance, to pay beyond payment card limits, or be immediately informed by the banks about a new payment request addressed to them. Request to Pay solutions have seen an increase of about 20 per cent in conversion rates at the checkout or an additional 25 per cent on the average transaction value for certain kinds of goods and industries, such as e-commerce, due to the aforementioned key benefits. that today the VoP is at the core of the payment journey. DSM As VoP continues to evolve, what additional features is CBI developing to further strengthen payment security for banks, corporates and citizens inside and outside of Europe? AC The Instant Payments Regulation is widely adopted in Europe, but in other countries there are other kinds of VoP schemes. In Europe, the VoP functionalities are connected to instant payment and traditional credit transfers – cross-border interoperability is very useful for all actors involved in payments. CBI is also working on other value-added services to be offered together with the Name Check for anti-fraud purposes, which could rely on data provided by third parties, to obtain a very detailed final risk score for both the payee and the payer. DSM What challenges does Request to Pay solve for corporates, citizens and public services? And how does it improve processes like invoicing, reconciliation and liquidity management? AC Request to Pay marks a milestone in the European payments landscape. It enables payment service users – corporates or final FFNEWS.COM

Request to Pay solutions have seen an increase of about 20 per cent in conversion rates at the checkout or an additional 25 per cent on the average transaction value DSM What are the early use cases and behaviours that suggest strong adoption potential for Request to Pay across the B2B and B2G landscapes? AC The first use case in Italy has been a B2G payment scheme involving public administrations. Thanks to the interoperability between our RTP solution and the public administration platform, we are able to orchestrate payment requests issued by public entities to citizens and businesses. We have a centralised system, which allows us to reach many of the public administrations in Italy, and this has been a vital strategic point. We have reached a wider public customer base, thanks to the public administration’s adoption of Request to Pay. The second step now is corporate use cases, and the initial market signals are

encouraging, because corporates are strongly requesting this solution. It could help them improve their efficiency in treasury management, in activities connected to the reconciliation of payments, and even allow them to offer innovative payment solutions capable of improving sales volumes and increasing the average transaction value. DSM Do you think there are any use cases where Request to Pay could be particularly useful, and have you come across any barriers to adoption? AC We definitely think some industries could adopt and exploit this solution better, particularly those that have frequent interactions with their final customers and that suffer from end-user spending limits, such as, for instance, the insurance industry and the telco industry, ecommerce and travel industries. A corporate could be unfamiliar with the technical aspects behind Request to Pay. But the payment service providers, together with system integrators, can smooth that process for them with a very simple ‘plug-and-play’ integration, so there really needn’t be any barrier to adoption. DSM Looking ahead, how do you see Request to Pay evolving across Europe and beyond? And what new functionalities could support more complex payment workflows in government and wider organisational contexts? AC Today, in Italy and Europe, we have a common scheme issued by the European Payments Council, which standardises all the messages that corporates and payment service providers have to exchange within Europe to send the Request to Pay. But interoperability beyond Europe would also be of interest to international business. A second innovation is collaboration-driven. Together with other financial entities, we could develop new services associated with Request to Pay. Other evolutions will be driven by technology. In Europe, for instance, we have the digital euro, which will be another important milestone within the payment industry, bringing technological changes. Request to Pay should also serve these kinds of payments for corporates and final users. For PSPs and corporates, Request to Pay is a very important business tool, especially if, as the numbers show, they use it to offer an innovative solution capable of bringing benefits to all the actors involved within the payment journey, no one excluded. DISCOVERSIBOS2026 9

DISCOV

S6 O SM IIAB 2 0 2 I M

THE AI PLAYBOOK

ARTIFICIAL INTELLIGENCE

As with any transformative technology, AI creates as many challenges as it solves. Red Hat is a global open-source enterprise solutions provider that has been working on tackling some of the biggest issues surfacing right now – namely, how best to integrate gen AI with existing architectures; how to make sure agents don’t go rogue; and how to reduce the big artificial footprint that AI could leave on the real word. We assembled a wide-ranging panel of Red Hat experts to get their take on all three

A flexible infrastructure for AI Open up the technology stack of a typical bank and you’ll find something closer to an archaeological dig than a tidy data centre. Mainframes at the bottom, a layer of virtual machines above, Cloud infrastructure on top and, increasingly, a fresh deposit of containers. Sitting atop many companies’ stacks, a brand new AI layer is now growing. The challenge isn’t simply where to put it, but how to make it get along with decades of technology sediment underneath. “The rate of change that’s come, particularly with AI, has exposed some real challenges,” says Ivan Jennings, Automation Platform Lead, UKI & CENE at Red Hat. The mainframes underpinning many UK and Irish financial institutions remain famously dependable, he says. But the journey into Cloud, Kubernetes and now AI has made ‘the speed at which institutions can get change into the market’ a far tougher proposition. For Dario Molinari, Senior Solutions Architect at Red Hat, the squeeze is tightest for institutions still leaning on ageing kit. Modernisation can mean ‘injecting new technologies like AI into end-of-life systems infrastructure’, he says. “That is the big challenge – just scaling the level of effort with ever-reducing budgets.”

10 DISCOVERSIBOS2026

The good news? Modernisation doesn’t have to mean bulldozing the dig site and starting again. Rather than babysitting each generation of technology separately, Red Hat argues for greater standardisation across the estate. Monica Sasso, Digital Transformation Lead, Global Financial Services, at Red Hat, says a common platform can act as a consistent layer across old and new technologies alike, taking the sting out of managing them. Crucially, standardising doesn’t mean modernising everything in one heroic undertaking. Sasso argues that firms should be able to move at their own pace and within their risk tolerance, without disrupting

practice. It lets virtual machines and containerised applications live side by side on a common platform, while also supporting AI workloads. That combination matters as financial institutions look to bring in AI without abandoning the applications they already rely on. Ashesh Badani, Senior Vice President and Chief Product Officer at Red Hat, describes the opportunity as getting mission-critical legacy applications and newer Cloud-native ones onto the same platform, then extending that foundation to AI models and agentic workflows. It may be best to think of this as progressive modernisation, rather than rip-and-replace. Existing applications keep running while

Injecting new technologies like AI into end-of-life systems infrastructure [is] the big challenge – just scaling the level of effort with ever-reducing budgets Dario Molinari, Senior Solutions Architect at Red Hat important business services. The idea is to abstract away some of the underlying complexity, freeing attention for the things that matter: security and innovation. Her prescription fits on a post-it note: standardise, simplify and automate. Red Hat OpenShift, the company’s Kubernetes-based hybrid Cloud application platform, is one way to put that into

institutions modernise individual workloads according to their own priorities, resources and appetite for risk. Standardising the platform, however, is only half the battle. The other half is identifying where all these workloads should actually run. For Badani, modern enterprises are increasingly hybrid by necessity, with workloads spread between their own data

FFNEWS.COM

centres and public Cloud. In financial services, those decisions come with extra homework: regulation, compliance, data restrictions and digital sovereignty requirements. Red Hat’s answer is what it calls open hybrid Cloud: using open-source technology to create a consistent foundation across public and private Clouds and on-premise infrastructure. Jennings puts the principle simply: “Open hybrid Cloud is all about being able to move your data, your applications, your tooling from one set of infrastructure to another, whether that’s in your own data centre or in the Cloud.” For financial institutions, that portability means more say over where different workloads live. It also ties into another pillar of Red Hat’s philosophy: open source. Jennings argues that openness gives organisations more freedom and reduces the risk of becoming beholden to a single vendor. Of course, freedom comes with its own paperwork. Running applications across different technologies and environments multiplies the number of processes

institutions need to manage consistently. Richard Harmon, VP & Global Head of Financial Services at Red Hat, points to patching, architecture modernisation and moving workloads between environments as prime candidates for automation. He describes it as ‘a foundation for any kind of architecture, hybrid in particular’. Red Hat’s Ansible Automation Platform is designed to automate IT operations across an organisation’s technology estate, including virtualised infrastructure, operating systems and networks spanning data centre, Cloud and edge. In financial services, that can stretch to operational resilience and disaster recovery, too. For Sasso, part of the value lies in making resilience boringly routine. She describes customers that, every day after trading closes, use automation to wipe their servers, rebuild them and restart them ready for the next trading day. If a failure or attack does strike, recovery isn’t a panicked scramble;

it’s something the team has already rehearsed hundreds of times. Rather than treating resilience as a tick-box activity, Sasso argues that automated playbooks can build it into the organisation at its core. The same logic applies to responding to failures and disaster recovery, including shifting environments when disruption hits. So, building for AI is less about tearing down what exists and more about creating infrastructure that can evolve alongside it. Standardisation tames complexity; hybrid architecture offers flexibility over where workloads run; and automation keeps an increasingly sprawling estate manageable and resilient. And AI won’t be sitting still. Harmon expects the technology and algorithms behind it to look very different within just a few years. Building for flexibility now isn’t only about accommodating today’s AI workloads. It’s about being ready for whatever turns up next.

AI on the inside: Not tearing down what exists, but evolving alongside it

FFNEWS.COM

DISCOVERSIBOS2026 11

Guardian or predator? AI defences Generative AI is growing up fast. Having mastered producing information, it’s evolving into agentic systems that can size up a situation, make a decision and act on it. For an industry trusted with customers’ money and sensitive data, that creates a double challenge: making sure AI systems behave as intended, and protecting them and the infrastructure around them from outside threats. Recent AI security headlines, including the Claude Mythos incidents and Hugging Face, have thrown those risks into sharp relief and divided the industry between those who believe agents are best kept corralled behind proprietary walls and those, like open source specialists Red Hat, who argue that transparency and shared responsibility are the only way to keep them in check. Any organisation seeing the swarms, breakouts and hacks of recent months is forgiven for feeling queasy – especially if it’s a financial institution. How should it keep increasingly capable AI secure, accountable and under control while the technology keeps moving?

For Richard Harmon, VP & Global Head of Financial Services at Red Hat, control starts well before an AI system goes live. Borrowing the software development principle of ‘shift left’, the practice of moving tasks like testing and security earlier in the software development lifecycle, he argues that institutions should set the rules of the road for AI before they start building anything. They also need to be able to see and question what AI systems are up to. Harmon points to strict controls, auditability and traceability, alongside rising demands for transparency and explainability.

Officer, EMEA, at Red Hat, puts it, doing the jobs expected of it and not wandering ‘outside of the domains that you are interested in’, which raises the obvious question of who is monitoring the situation. Ivan Jennings, Automation Platform Lead, UKI & CENE at Red Hat, envisages increasingly ‘self-healing infrastructure’, where AI fixes technical problems while a human stays in the loop, keeping an eye on things. Regulators are thinking along similar lines. The EU AI Act introduces transparency and risk-management requirements for certain AI systems, while the NIST AI

As organisations transition from experimental AI pilots to long-running autonomous agents, establishing clear operational guardrails becomes a critical infrastructure requirement Steven Huels, VP of AI Engineering at Red Hat “With AI, as it becomes more autonomous in terms of starting to evaluate systems and making decisions, physical decisions, the guardrails and the transparency and explainability are key," he says. For agentic AI, that means staying in its lane, or as Julio Guijarro, Chief Technology

Risk Management Framework gives organisations a structure for managing AI risks throughout their lifecycle. For financial institutions, the stakes are especially high: AI innovation has to coexist with regulatory obligations and the duty to protect customers’ money and data.

AI’s hidden bill Counting the cost of resources Banks have spent years scrutinising the environmental footprint of the companies they finance. As AI adoption grows, they may need to look a little closer to home: at the energy and water consumed by the technology humming away in their own operations. Red Hat argues that seeing where those resources go is the first step to using them more efficiently. Every model, however clever, has a physical address. AI training and deployment happen mainly in data centres, where servers draw electricity and cooling systems can consume both power and water. The

12 DISCOVERSIBOS2026

Hungry tech: Measurement is key to controlling AI’s impact

International Energy Agency estimates that data centres accounted for around 1.5 per cent of global electricity consumption in 2024 and expects their electricity demand to more than double by 2030, with AI the biggest driver of that growth. The UK Environment Agency says large data centres can use millions of litres of water a day, and growing AI use is expected to push up demand for data-centre capacity. As financial

institutions roll out more AI, understanding the resources behind it matters not just for sustainability, but for efficiency and cost. You can’t cut what you can’t see. Julio Guijarro, EMEA Field CTO at Red Hat, argues that better data gives organisations visibility into which computing processes are eating energy, and where there’s room to trim. “For sustainability, and really any green

FFNEWS.COM

DISCOV

Hands on: The open source community is one layer of security

Governance can’t afford to wait for the technology, or the rulebook, to stop moving. Guardrails are one layer of defence, for Red Hat, open source offers another – toughening up the technology itself through transparency and collective scrutiny. Put simply, open source makes the underlying code available for anyone to inspect, test and contribute to. Red Hat argues that this lets a wider community of developers, researchers and organisations spot vulnerabilities and build safeguards together, rather than relying on a single technology provider to catch everything.

agenda, we need the data,” he says. One approach Red Hat has worked on is Kepler – Kubernetes-based Efficient Power Level Exporter, an open-source project that estimates and reports the energy consumption of processes, containers and Kubernetes pods. Think of it as a smart meter for your workloads: those metrics can then be used to spot and optimise energy use. For Guijarro, visibility can also change behaviour. Showing developers and users the energy impact of what they build gives them something to act on when deciding how to design and run it. “If there is no data, you don’t see what is happening, how the work that you are doing affects the consumption of technology – and then it’s very difficult to change it,” he says. A dashboard, however, is only useful if someone acts on it. Dario Molinari, Senior Solutions Architect at Red Hat, argues that infrastructure efficiency and sustainability are becoming hard to tell apart. He points to the shift towards leaner, containerised infrastructure as one way

FFNEWS.COM

ER S I B OS M IAM

Harmon sees that many-eyes model becoming increasingly important for AI: “The goal here is a global community ensuring everything is secure and safe,” he says. “It’s not one single firm who owns a monopoly on a particular capability.” Guijarro argues that meaningful openness means understanding how models were built, what data went into them and which tools were used in order to trace how decisions are made. Red Hat is applying that open-source thinking to AI’s emerging security challenges. In July 2026, the company became an inaugural member of NVIDIA's Open Secure AI Alliance, an initiative focussed on developing open tools and techniques to protect the AI stack, from open-weight models all the way through to agent harnesses. Writing principles and regulations down is one thing. Turning them into controls that actually run alongside live AI systems is quite another. As Steven Huels, VP of AI Engineering at Red Hat, puts it: “As organisations transition from experimental AI pilots to long-running, autonomous agents, establishing clear operational guardrails becomes a critical infrastructure requirement.” Closing that gap is the focus of asago – AI Safety And Governance Orchestration, a collaborative open-source community

I 2026 project that Red Hat announced in August. Asago aims to translate corporate and regulatory AI governance policies into working operational controls. Importantly, the project is designed to plug into frameworks that institutions may already be wrestling with, including the NIST AI Risk Management Framework, OWASP LLM Top 10 and the EU AI Act. And it isn't a Red Hat solo act: founding participants include Microsoft, NVIDIA, IBM Research, MIT Lincoln Laboratory and The Alan Turing Institute. Asago is still in its formation phase, but its premise speaks to the need for governance to start moving at something closer to the speed of the technology as AI systems become more autonomous. For financial institutions, the goal isn't to pick between AI innovation and safety. It's to build an environment where the two can advance together. As AI takes on more autonomy, financial institutions will need clear guardrails around what it’s allowed to do. Red Hat sees open-source transparency and collective scrutiny as one line of defence, alongside governance that can turn policy into practical controls. The technology won’t stand still, and neither can the systems built to keep it honest.

financial institutions have slimmed down the physical resources needed to run applications, while better monitoring, metering and telemetry show developers exactly how much an application consumes. For Molinari, that makes sustainability as much a question of economics as of environmental responsibility. Organisations that don’t prioritise it, he argues, could end up feeling it twice: in their cost base, and in the growing scarcity of energy and other resources.

with take sustainability seriously because they have ‘a duty of sustainability towards their investors’. Many also offer sustainability-focussed products, which gives them another good reason to check how efficiently their own IT runs. For financial institutions, getting a grip on AI’s growing appetite starts with understanding it. Better visibility shows where energy is being consumed, while more efficient infrastructure can cut both resource use and cost. And for institutions asking the

If there is no data, you don’t see what is happening, how the work that you are doing affects the consumption of technology – and then it’s very difficult to change it Julio Guijarro, EMEA Field CTO at Red Hat There’s one more consideration for financial institutions: what they expect from the businesses they finance may increasingly need to show up in their own technology operations. Practising what you preach, in other words. Molinari says the banks he works

businesses they finance to meet sustainability commitments, their own technology estate is increasingly part of that conversation. As AI scales, how lightly it treads on the planet’s resources may become one of the most important measures of its success.

DISCOVERSIBOS2026 13

CONVERGING CO-EXIST TENCE SECURITIES

When it comes to tokenised securities and other digital assets, a common view is emerging: they need to work inside the system we’ve got, not run in a parallel universe. We spoke to Euroclear’s Isabelle Delorme about cohesion, co-existence, and collaboration

It’s an interesting time for the world’s central securities depositories (CSDs). They are seeking to manage a shift towards co-existence and convergence, where traditional and digital assets can come together under unified operational and legal frameworks. Rather than plotting how to replace legacy finance, CSDs are acting as a fundamental bridge between traditional finance (or TradFi, as the hipsters call it) and the new wave of digital and programmable money, whether that’s stablecoins, tokenised deposits, or real-world assets (RWAs). Financial market infrastructure group and CSD Euroclear stands at the forefront of this convergence, processing and finalising trades and providing custody for trillions of euros. Isabelle Delorme is the Global Head of Product Strategy & Innovation at Euroclear, tasked with bringing cohesion across liquidity products, fixed income and equity, and funds and ETFs in three geographies – the Americas, Asia, and Europe. “There’s a lot on my desk – or on my screen,” she says. “Strategy papers, regulatory reports, client feedback, employee feedback, and everything around how we can service our clients while modernising the markets. It’s 14 DISCOVERSIBOS2026

hectic, fascinating, and exciting – but the driver for me is the diversity, and the possibility to have a broad perspective on what the market wants and where we all go together.”

Creating new connections

with other technologies,” says Delorme. “It’s important we equip ourselves with the right resources to make sure that we can deliver for our clients at scale in the short term – not in 10 years.”

Euroclear has long held the view that solving A maturing view on new possibilities liquidity for digital securities rests on It’s fair to say Euroclear is something of connecting digital and traditional markets, a pioneer in the digital assets world. For rather than attempting to build entirely example, it facilitated the first live issuance new liquidity pools. of a digital bond in October 2023, supported True adoption requires interoperability, by R3’s enterprise DLT platform Corda. continuity, and access to the deep trust The company is now building up its own networks that already support global capital internal capabilities, and its approach markets, which have been built up over reflects the wider, maturing view of DLT and decades. There have always been specific tokenisation illustrated by new initiatives settlement assets for specific use cases. that are investing heavily in the space. “Different monies solve different problems,” This includes Project Agorá, an expansive explains Delorme. “Central bank money is public-private collaboration comprising of important for trust and sovereignty, but many of the world’s major central banks and commercial money is everywhere – it’s what financial institutions, which is focussed on we use for cross-border settlement and for converging the underlying infrastructure of many applications. wholesale cross-border payments using “If we translate that to digital money, it’s tokenisation and the correspondent banking going to be the same, so we will need a variety system. There is a growing realisation that of settlement assets. What really matters is ‘old’ and ‘new’ money will need to coexist for interoperability, because nobody wants to a long time – maybe forever. create fragmented liquidity pools. We need to As far as the use of DLT in securities make sure that everything can co-exist across settlements goes, the Eurosystem took a one global system.” major step forward in March 2026 by Euroclear’s International Central allowing DLT-based assets to What’s Securities Depository (ICSD) – be accepted as collateral, missing is a clear Euroclear Bank – now interacts so long as they are issued idea of how we’re all with digital assets through its via DLT-based CSDs, settled going to get returns on proprietary Digital Financial in systems connected to the considerable Market Infrastructure (D-FMI), TARGET2-Securities (the investments we’re making to leverage these blending distributed ledger Eurosystem’s single technical technologies technology (DLT) with platform for central bank at scale. traditional core settlement money securities settlement), systems. This helps to bring together and meet standard Eurosystem risk the best of both domains – combining and management criteria. DLT-enabled efficiencies with the resilience, This decision highlights the Eurosystem’s governance, and liquidity frameworks of continued commitment to encouraging traditional market infrastructure. innovation and technological progress, “We’re building an organisation where we enhancing market efficiency, and have dedicated people working on the contributing to the future integration of transition to DLT, and enabling it to interact European capital markets. FFNEWS.COM

DISCOV

ER S I B OS M I A M I 202

In a statement issued following this announcement, Euroclear said the development places digitally issued securities on the same footing as traditional ones, and is a meaningful milestone for the entire market, a clear signal that digital assets are being integrated into the core of Europe’s financial infrastructure. For Euroclear, the Eurosystem announcement is much more than a policy update; it’s validation of years of purposeful investment, innovation strategy and industry collaboration. “What’s changed is that we now have clarity on common expectations,” explains Delorme. “After years of experiments, and years of speeches, the collective aim is becoming much clearer.” While Euroclear and other companies have cracked the technical viability of DLT, the question now turns to how they can unlock commercial viability. “We’re implementing these new technologies, but which financial goals do we have in mind?” says Delorme. “Because that’s probably what’s missing – a clear idea of how we’re all going to get returns on the considerable investments we’re making to leverage these technologies at scale. I think Sibos provides an excellent opportunity to discuss this aspect.” This question perhaps helps to explain why it’s taken so long for the industry to get to grips with DLT. After all, it’s been on the blocks for around 20 years with a long line of ‘test transactions’, ‘proof of theory’, and FFNEWS.COM

commercial partnerships – like those with R3 Corda, which is rumoured to still not be profitable itself.

Client-centric developments While discussions around commercial viability, regulation, and integration rumble on in the background, the only thing that Euroclear’s clients are concerned about is the speed and accessibility of the trade. “We know that nobody wants more fragmentation,” says Delorme. “But what do they want? Firstly, I would say the typical capital markets user wants to ensure their ability to move assets and collateral across markets, across networks, and now across ledgers. “They also want less operational friction, and this is what technology can bring. And then thirdly, they want true interoperability between technologies, use cases, and platforms. “We shouldn’t be techno-evangelists. We’re driving infrastructures, and we are there to ensure that behind these three elements we still keep in mind that what matters the most is trust, resilience, and liquidity efficiency.” Delorme is pragmatic when it comes to how Euroclear is adopting other emerging technologies. This includes AI for which the starting point has to be good data hygiene, she says. “In capital markets, we’re sitting on massive volumes of data. So a big focus for our investment at present is around

data quality and data governance – and then we can apply the AI to extract additional benefits. “The most exciting initiatives aren’t always glamorous. They’re about operational efficiency and using AI for better controls. It’s about making sure that we are able to automate a lot of the processes that were previously manual and fragmented.” Looking ahead, Delorme puts an emphasis on collaboration and partnerships in designing the future. “I don’t think our clients come with a clear expectation of what CSDs should become or what they should provide around digital assets,” she says. “That’s what makes it very interesting. We have to co-create it together.” This point is well illustrated by how the conversation around tokenising fixed-income securities has changed. “Tokenising fixed income felt like bread and butter and not that fancy a use case,” explains Delorme. “We were considering whether we should change focus on asset classes. But, after a series of conversations at Sibos in Frankfurt last year, it became clear to me that we have enough volume to drive the market to something that justifies a collective ownership and a collective acceleration. Something that can drive the transition.” It will be interesting to see what shifts the dial in Miami.

Isabelle Delorme, Global Head of Product Strategy & Innovation at Euroclear

DISCOVERSIBOS2026 15

London 8 October, 2026

Secure your spot: red.ht/london26

DISCOV

ER S I B OS M I A M I 202

The ˊnewˋ

CROSS-BORDER

correspondent banking system Tokenisation and DLT are not incompatible with using the established network of intermediary banks in wholesale cross-border settlement. And Deutsche Bank very much has a foot in both camps

When Deutsche Bank completed a treasury payment as part of Project Agorá this summer, the media coverage focussed on the tokenisation of cross-border wholesale payments and how the project had achieved all-or-nothing atomic settlement. Deutsche Bank was the intermediary agent for a €10,000 transaction executed on chain between Lloyds Banking Group and CaixaBank, one of 17 payments that tested the ability of a multi-layered distributed ledger technology (DLT) to turn sequential cross-border processes into simultaneous ones. FFNEWS.COM

The average transaction time was an impressive 80 seconds from start to finish. But what was perhaps less talked about was that Project Agorá is a hybrid solution to the well-known problems associated with the traditional correspondent banking system – cost, time and opacity. It’s a blueprint for a stepping stone approach to next-gen settlement, designed to capture the structural benefits of DLT without shutting out the existing two-tier global banking framework. What it proved was that it is possible to substantially accelerate settlement on a shared ledger while still preserving jurisdictional autonomy for the banks – and, no, it doesn’t require new regulatory frameworks to do it. In other words, the old and new worlds can co-exist but with dynamic advantages. Such a pragmatic response could ultimately advance the broader case for tokenisation, specifically in wholesale trade where multiple actors (businesses, lawyers, agents and banks) must be convinced, corralled and co-ordinated. Deutsche Bank itself doesn’t pin its future global trading strategy on

Tsvetanka Nankova, Global Head of Sales for Institutional Cash and Trade Finance at Deutsche Bank

a single system, either, but, given the almost 20 years of global experimentation around tokenisation and DLT, Tsvetanka Nankova, Deutsche Bank’s Global Head of Sales for Institutional Cash and Trade Finance, believes banks have now got to get to grips with it. “As institutions, we’ve been experimenting across multiple areas,” she says. “The question now is how do we take that and bring scale to some of these initiatives?” She describes correspondent banking as ‘becoming a truly intelligent, always-on digital business across multiple rails’, and she’ll be keen to advance that view among peers at this autumn’s Sibos conference in Miami. Deutsche Bank began investigating the potential for DLT as an international payment solution way back in 2014. The bank remains an active member of the R3 Corda Network (which is a likely contender to become the permanent platform for Agorá) and last year it completed its first euro-denominated cross-border payment on the Singapore-based fintech infrastructure provider Partior’s blockchain. DISCOVERSIBOS2026 17

At the same time it continues to improve the traditional rails of correspondent banking. For example, this year it has developed the AutoConvert feature of its FX4Cash platform, which reduces settlement times by automating foreign exchange handling. “Transactional FX is one of the areas that’s going to deliver additional revenue streams for financial institutions such as Deutsche Bank, so it’s a key focus for us,” says Nankova.

The role of AI While FX4Cash relies on a highly automated rules-based system, Nankova believes AI is going to be central to banking’s shift from being a ‘volume and balance sheet-based industry to a data and digital solutions industry’. But banks, even banks as big as Deutsche Bank, need to be highly focussed on where it can deliver them a competitive advantage. She says: “What is important is that AI and data are not simply driving the development of very fancy dashboards. They should really be driving opportunities across cashflow optimisation, payments routing, liquidity management and so on. There are lots of ways it can drive improvement for our clients and also generate new revenue flows.” Specific areas where Nankova believes AI can make a crucial difference are operations, transaction monitoring and cash flows.

Balance sheet and risk intermediation is an area where new entrants in our space will struggle to scale. Having that regulated balance sheet is extremely important, especially in times of turmoil

Tsvetanka Nankova, Deutsche Bank “Operations teams handle a lot of inquiries and investigations, and at Deutsche Bank we already use AI to ensure inquiries we receive are routed to the right partners or colleagues, so they are resolved professionally and swiftly,” she says. “Thanks to ISO 20022 and the improvements it has brought to data, we’re seeing a significant reduction in the number of alerts generated by our transaction monitoring systems. By using AI, we can swiftly clear the false positives and focus on the alerts that matter. “Specifically in the revenue-generating space, AI can be used to forecast cashflows, 18 DISCOVERSIBOS2026

which allows us to schedule payments on behalf of clients more efficiently. Treasurers at corporates and financial institutions are absolutely looking for the optimisation of liquidity management as the world becomes more complex and fragmented. “They are also looking for swifter decision-making and AI can enable us to take better and faster decisions in the credit risk management space. But it is crucial to ensure that our models are extremely well tested. “Banking is built on trust and our management, regulators, and clients have zero tolerance for errors. One mistake can absolutely take you out of business or have significant repercussions on everything you do. So, the way we implement artificial intelligence

going to be critical, and ensuring that we are really concentrating on the areas where we have a competitive advantage rather than doing everything. “In the past, I think the mindset across financial institutions was like, ‘oh, this is new, right? Sounds a bit sexy. Let’s look at that!’ And, you know, they maybe added a little bit of tactical value short term, but they were not necessarily strategic to your institution.” In terms of where a major bank’s advantages lie, Nankova argues that boardrooms need to remember that size can matter. “Balance sheet and risk intermediation is an area where new entrants in our space will struggle to scale,” she says. “Having that

Data advantage: AI is helping banks turn information into faster, smarter decisions

matters. It must be done with responsible velocity and by making sure that we still have a human in the loop.” Despite AI’s capability, she believes a blanket approach to implementing the technology is the last thing banks should adopt.

Be more fintech Nankova has spoken in the past about the fintech advantage and the agility of monoline service providers, and says being ‘more like a tech company’ is key for financial industry incumbents if they are to respond as quickly to customers’ and regulators’ demands as the sector’s newest entrants do. “But we need to understand, and it’s a shift in mindset, that we cannot own every single touch point with our clients. We need to be extremely smart about what we do because there are only a few strategic bets that we can invest in,” she says. “They’re costly, but you also need the time to focus on these. Ultimately, ruthlessly prioritising what we’re focussing on is

regulated balance sheet is extremely important, especially in times of turmoil. That’s when clients are less worried about whether a [technology] ecosystem looks fancy or not. They’re more worried about who’s going to provide the liquidity that they need, who’s going to provide the backing to meet their financial obligations. “Another area where I think we as financial institutions have been strong, is that execution at scale through a lot of rich operational data that we sit on – and certainly new technologies such as artificial intelligence can employ that going forward.” Being focussed on the task at hand – namely improving the trade finance machine and specifically cross-border payments – while not being committed to one technology, is how Nankova believes Deutsche Bank will succeed in serving its customers. “Ensuring that we’re investing in strategic areas – areas where we can absolutely have a competitive advantage – is going to be extremely important.” FFNEWS.COM

DISCOV

S6 O SM IIAB 2 0 2 I M

Navigating PAYMENTS

The world can feel like a very small and unpredictable place if you’re a business engaged in global trade. So how helpful is technology in steering them through troubled waters? While news around disruption in the Strait of Hormuz focusses on fuel prices, choking off 20 per cent of the world’s oil is also a trade and banking story. Industries such as transport, agriculture and manufacturing felt 20 DISCOVERSIBOS2026

a squeeze on margins within weeks of restrictions being placed on shipping earlier this year, and resulting liquidity pressures increased fast. Add in tariff warfare, sanctions regimes and private credit strain, and it’s clear that 2026 has been a torrid year for trade finance. So, what does this mean for those tasked with keeping businesses and trade afloat? “When I talk to treasurers, they’re facing geopolitical risk, liquidity risk and foreign exchange risk,” says Heather Crowley of J.P. Morgan Payments. “In the global environment, people are currently working to protect themselves. Whereas treasurers used to be focussed on cash from a working capital perspective, now they are also consumed by risk and resilience. They’re managing a lot at the same time.”

Heather Crowley, Head of Supply Chain Finance, Core Trade Product and Product Delivery at J.P. Morgan Payments

Embedded finance Crowley’s job as Head of Supply Chain Finance, Core Trade Product and Product Delivery is to support the bank’s clients through these real-time events, while defining J.P. Morgan Payments’ product roadmap, so businesses can achieve and maintain the liquidity that’s crucial to navigate change. The key theme right now is frictionless trade finance products, she says, which embed and integrate banking services into a client’s own enterprise resource planning system. J.P. Morgan Payments is a big player in this space. In April it launched the Working Capital Accelerator platform that centralises the bank’s working capital products, and in August won the Working Capital Innovation Award from The Working Capital Forum for its Supply Chain FFNEWS.COM

Finance solution, hosted within the Oracle Fusion Cloud ERP. Crowley says: “The tools clients are asking me for are still, for example, supply chain finance or receivables, but how I deliver them is changing. “You see a lot of native solutions built into ERPs, whether that be SAP or Oracle, where the client can just turn on these solutions without having to invest in [extra] tech to bring it about – it’s frictionless at their end. The onus is on us as providers to take on this work. “Clients also want the process to be frictionless for their suppliers, all the way down to their small and medium-sized suppliers. Tools have to be flexible as well. What the last five years has shown us is that the supply chain can change overnight. So the tools we deploy must be highly flexible, global, yet meet local needs.”

Pragmatism in practice The increased pressures on treasurers brought about by geopolitical instability and change mean the pursuit of speed and efficiency is vital. Crowley explains that at the beginning of the 2020s, platforms were launched that offered every trade finance instrument in one place, but banks failed to identify the incentive for using them. Six years on, Crowley is hopeful AI will now improve digital integration on a more practical level. “We saw a lot of these utopian solutions, all the trade instruments on one platform, and [the idea was that] the buyer, seller, the buyer’s bank, the seller’s bank, the freight forwarder and insurer would all go to the site to use them,” she says. “It didn’t happen. It was too large to take on, so it never took off. “As a trade organisation, what we’ve seen are more pragmatic solutions, where, for example, the eBoE [electronic bill of exchange] has a meaningful impact for the client. “I do think, with AI being added, we’re going to see interoperability, and [such platforms] will take off. We were too euphoric about the whole thing before. We have a lot of fintechs in the fintech graveyard, but the ideas were sound. “I always tell my clients: pick an instrument you already do today. Pick a provider, a bank that you’re comfortable with today. Digitise that piece that’s meaningful to you and to your industry. If we all do those pieces externally, that interoperability is going to come together.” Crowley says that in the race to create AI-powered solutions, banks and fintechs must focus on what matters to the client. FFNEWS.COM

“They need frictionless, seamless, integrated,” she explains. “How we get there is less important to them. I don’t think what I’m being asked for from my clients necessarily changes. Even if we start thinking about a future state with a client’s AI speaking to my agents etc, what they’re really going after is the experience. “They want it faster, they want it specific to their needs. It’s all about how we deliver the experience, and I think AI is going to allow us to provide that in a better way.” Crowley adds that she has seen AI deliver the ‘biggest bang’ where it can be put to work in a particularly complex area. “So, for example, within my back office, this is where I can control four million pieces of paper coming into my shop. If I can take pieces of that and I can utilise the AI tools, then I can really make a difference from a trade perspective. “And if you think about trade finance, it’s not just the paper that slows us down. It’s all the regulations that we have to work on. Remember the Dear CEO letter out of the EMEA and so on – that’s what slows us down. And that’s a high-cost element.

What the last five years has shown us is that the supply chain can change overnight. So the tools we deploy must be highly flexible, global, yet meet local needs Heather Crowley, J.P. Morgan Chase “If we can take those elements and apply tools such as AI to them, I can make a meaningful difference within J.P. Morgan Payments’ back office and deliver bottom-line impact to trade finance. “When you’re doing an individual instrument for clients, that’s a slower take-up. But if the banks start digitising amongst themselves, the clients start digitising with the instruments, we will truly drive change. There’ll just be less noise because we’re going to be driving that holistically.”

Taking a step back Once armed with an effective suite of trade finance tools, Crowley says executives faced with a working capital challenge must step back and think holistically. That’s the advice she

shares when she sits down with chief financial officers and treasurers. “We should all remember that these challenges sometimes come as the result of an event. Never react defensively to an event, a single data point or a single indicator because it might cause damage at another element of your cash conversion cycle. “Working capital is just one aspect where you can build resilience against some of the events out there, such as supply chain vulnerabilities or interest rate increases. “So, don’t jump right into a working capital conversation. Look at it holistically. Look at your liquidity – that will help you identify your working capital gap, and help you understand what else you need to focus on. “Also, look at your peers in your industry and identify where there might be outliers. And with all these artefacts – your objectives, your holistic liquidity position as well as your working capital, and what else is important to you in this calendar year or time period – we [the bank] can look at what is the optimal working capital solution for you.”

Eye on the future Advising a client effectively means knowing their business, and Crowley continually encourages her staff at J.P. Morgan Payments to meet and work alongside clients so they can foresee problems that will need solving before they cause any chaos. “You have to be futuristic, constantly thinking about what’s coming,” she says. “Things change from day to day and clients’ needs change. And if you want to be a trusted adviser, if you really want to put something on the product roadmap, it takes a while to develop it. “When associates and analysts come into J.P. Morgan Payments, I tell them they need to know their clients, be valuable and constantly be prepared to learn from them. “And I always encourage them to learn the business from the bottom up. Really get to know how you implement a programme, and ask what are the key pieces from a technology perspective? “I tell the entry-level staff: get right in there. I want you to make those phone calls. I want you to onboard a supplier. I want you to implement a programme. “Because 15 years from now, you’re not going to have time to do it. So, take the time when you come in and learn the details. That way you become very valuable to a client.” DISCOVERSIBOS2026 21

DISCOV

S6 O SM IIAB 2 0 2 I M

At some point, you have to stop testing the future and start using it. Tokenisation has generated plenty of proofs of concept, controlled experiments and ambitious predictions about the capital markets. BNP Paribas’ Securities Services business has been involved in many of them. Now it’s discovering what happens next. In 2025, BNP Paribas Asset Management worked with Allfunds Blockchain and BNP Paribas’ Securities Services business on natively tokenised money-market fund shares and cross-border transactions between Luxembourg and France. More recently, the bank has used public blockchain infrastructure to issue a tokenised share class of a French money-market fund on Ethereum within a controlled framework. That’s against a background of industry initiatives such as the Bank for International Settlements’ new Project Agorá, which is exploring how tokenised commercial bank deposits and central bank money could transform settlement. For Wayne Hughes, Digital Assets Market & Client Engagement Lead, Securities Services, BNP Paribas, the most revealing lessons have come from moving beyond the safety of the sandbox. And he says, once tokenisation enters the real world, it turns out the technology isn’t necessarily the hardest part. DISCOVER SIBOS MAGAZINE Tokenisation has been talked about for a long time. What has changed over the past 12 to 18 months to move the industry from experimentation towards production? WAYNE HUGHES We all needed that first phase in order to start getting comfortable with this technology and what it could mean

for our industry going forward. Through that experimentation, we were able to at least start to address a lot of the challenges that were holding back potential scalability. Now, that experimentation has started to pay off. We also have a lot more regulatory clarity and there are some quite large, visible initiatives in the market now that are pulling together an ecosystem, which was one of the challenges before. From our perspective, it might be some time before we get to significant volumes, but the last couple of years have been more focussed on learning by doing. We’ve been experimenting in a live environment, both to test our technical foundations and to begin engaging with our clients and market counterparts. I definitely feel that we’re in a new stage of our industry’s digital asset journey DSM BNP Paribas’ Securities Services business has now been involved in live initiatives with Allfunds Blockchain as well as work involving tokenised money-market funds on public Ethereum. What did going live teach you that a proof of concept could not? WH The proofs of concept were great to get the ball rolling and test out the technology, but you could argue that, at some point, they were too easy. Just by their nature, the scope was very limited, and they didn’t really give us that opportunity to get the full organisation engaged. That really changed when we commenced live experimentation and all the different areas of the bank started to see what this could mean for them going forward. A clear example was our first initiative – the first time we signed a contract with clients. That process, even for this very simple experiment, took quite a few months and was clearly related to the expertise of the lawyers

involved. The first question I got from them was: “What is a blockchain, and what does it mean for me from a legal perspective?” That’s what these live experiments really did for us. They allowed us to get the full organisation engaged, start to get people up to speed and comfortable with this technology. You could also argue that it gave them an opportunity to raise some healthy concerns and for us to start working through that. DSM Did the live environment also change your understanding of where the biggest risks and difficulties actually lie? WH When we first started work on this there were a lot of concerns about the technology, about the risk it could potentially introduce and topics around scalability. This wasn’t where the concerns were concentrated when we went into a production environment. They were much more around legal and operational topics. That is part of the value of doing these things live. Earlier on, much of our work was about laying the technical foundations for what we believed would become our future services – everything around connectivity to distinct blockchains, wallet management, tokenisation and compliance. Now the day-to-day is a combination of executing on our initial product roadmap while continuing to explore some of the newer or less mature use cases. Most importantly, it's about engaging with clients to understand what the requirements are going to be in future and making sure we’re ready to cover those requirements. DSM What are clients actually asking the Securities Services business for today? Has demand changed as tokenisation has moved closer to production? WH Right now, the noise in the market

The future is DLT

here (almost)

BNP Paribas’ Securities Services business has moved beyond the sandbox and kicked the tyres of tokenisation in the real world. What has it taught the team so far? 22 DISCOVERSIBOS2026

FFNEWS.COM

concerning use cases is especially around collateral mobility, which we believe is one of the promising use cases going forward. But for our existing client base, most of the demand we’re seeing is for tokenised money-market funds. And it’s mainly still a very straightforward distribution play, normally into their retail channels. So, real demand right now is mostly to do with the cash side – we see real interest for extended operating hours for cross-border corporate payments, for example. A lot more clients are ready to consider doing something in a live environment. Most of those are considering what could be their first experience in production. They’ve probably done a lot of proof of concepts, but they haven’t yet jumped into something in production.

I definitely feel that we’re in a new stage of our industry’s digital asset journey… the last couple of years have been more focussed on learning by doing

With that in mind, a lot of the demand is for relatively simple use cases, with the objective of having that first experience and then being able to build upon it. DSM As more institutions go into production, does interoperability become the next big obstacle? How do you avoid replacing today’s fragmented infrastructure with a new collection of blockchains, platforms and standards? WH That’s a key point, because we definitely run the risk of replicating the past, and maybe even making it worse in some ways, with another solution becoming another standard. The existing fragmentation is normal – it’s very early days. But consolidation and addressing that fragmentation will be critical going forward if we want this to scale. That said, we don’t necessarily believe the right target is one chain, one platform and one standard. A certain level of healthy competition is positive. There could definitely be different solutions that are better suited to different use cases or different regions. From our perspective, it’s more about seeing whether we can come up with a reasonable, limited number of chains, platforms and standards – not

replicating the past, but neither waiting for this perfect world to happen. What I believe is important is for the industry to collaborate, which has never been our strong point, maybe, in the post-trade world. If we want to address fragmentation and get to a point where there is true potential for scalability, we believe that will only happen if we collaborate and pull together an ecosystem around at least a limited number of chains, standards and platforms.

Wayne Hughes, Digital Assets Market & Client Engagement Lead, Securities Services, BNP Paribas

DSM After years of discussion and experimentation, what would convince you that tokenisation had finally become part of mainstream capital markets? WH It would be great if we were talking less about what tokenisation could provide and instead much more about the tangible value we’re seeing it deliver. There has been a lot of good work. We’ve got some benefits out of everything that has been done, but I really think we in the industry are ready to take that next step. We need to find a way to converge and create an ecosystem, and move towards a point where tokenisation becomes part of the normal function of capital markets, rather than a collection of individual initiatives.

Going live: Tokenisation is making the leap from experimentation to production FFNEWS.COM

DISCOVERSIBOS2026 23

Real-time visibility into every payment.

The platform built to answer “where is my payment?”

Booth L026

vyntra.com

DISCOV

ER S I B OS M

Upgrade the PAYMENT RAILS

I A M I 202

tech, but trust still ll moves the money Innovation will be everywhere at Sibos in Miami. But for Simon Eacott of NatWest, real progress happens where old meets new Ten years ago, it was easy to imagine that the payments revolution would sweep away much of the infrastructure on which international banking had been built. Blockchain would disintermediate banks, and new rails would render old ones obsolete. But a decade of payments innovation hasn’t killed the two-tier banking system or entirely displaced the established rails it uses, as many had predicted it would. For NatWest Head of Payments Simon Eacott, the really interesting transformation is happening as these old and new networks converge. “Correspondent banking is alive and kicking, it just looks and feels different,” says Eacott. As the global banking community gathers in Miami for Sibos 2026, that observation provides a useful counterweight to a conference appropriately themed Digital Finance For AI-Driven Economies. Because while the technologies occupying the industry have changed dramatically, one characteristic of payments has not. “Payments is a network business,” says Eacott. “It only operates if every link in the chain is working well together.” FFNEWS.COM

And Swift itself illustrates how those links are being re-engineered for the modern payments world. More than 11,000 financial institutions remain connected through its network, which has underpinned international payments for decades. Yet Swift is simultaneously evolving its infrastructure around richer data, enhanced cross-border payments and digital assets. NatWest is also participating in another example of pragmatic payments evolution: Project Agorá, the Bank for International Settlements-led experiment exploring how tokenised commercial bank deposits and central bank reserves could operate on a shared programmable platform. Agorá moved beyond theory in July. Twenty-eight financial institutions and central banks across Asia, Europe and North America completed real-value transactions totalling around CHF800,000 (Swiss francs) across 17 scenarios. Those included corporate and interbank cross-border payments, with an average initiation-to-settlement time of around 80 seconds. Participants also gained end-to-end visibility of payment status and routing. So, the reality may be less about blockchain replacing

Simon Eacott, Head of Payments at NatWest UK

correspondent banking than about technologies such as tokenisation making established, collaborative models work differently. For Eacott, evolving alongside traditional, trusted pathways rather than leaping over them is important. How that’s achieved is likely to be a top topic at Sibos where industry’s shiniest new things are on display. “It’s where the future of payments gets discussed, it gets stress tested, and it’s where big ideas meet the practical realities of delivery,” says Eacott.

Seeing the money That’s a reality check that resonates with corporate treasury. Instant payments attract headlines, but speed alone is not the whole of the value. The combination of 24/7 payment infrastructure, APIs and richer ISO 20022 data is changing what corporate treasurers know about their liquidity – and how quickly they can act on it. “If you look at it through the treasurer’s eyes, it’s the visibility of that cash in real time that’s most important,” he says. “The actual instant settlement behind it is the mechanics that the banks need to do.” DISCOVERSIBOS2026 25

For multinational businesses, that can mean managing liquidity as operations ‘follow the sun’, instead of waiting for overnight batches or weekends to pass or information to arrive from disparate banking systems. “I think the value is not about moving it faster,” Eacott says. “It’s about knowing where your money is at any one time and exactly what you can do with it.” Real-time payments can accelerate order-to-cash and working-capital cycles, while APIs can connect bank data directly with treasury management and enterprise resource planning systems, enabling faster reconciliation, forecasting and liquidity decisions. In other words, the payment is becoming as much about information as movement. At the other end of the corporate payments chain, open banking is creating new ways for businesses to collect money. The UK’s regulated open banking framework has created common standards for banks and third parties, while the US has historically developed through a more fragmented, market-led model. NatWest has put those UK rails to work through Payit, its open banking proposition. Businesses can collect money without storing a customer’s card details, with payments settling within seconds over Faster Payments. Payit also supports payouts and refunds, account validation and recurring payments, and integrates through APIs. For merchants, near-real-time settlement

That may be a recurring lesson of payments innovation: customers rarely care about the rail. They care about what it enables. But Sibos 2026 will address a new complication. What happens when the party initiating a transaction is no longer necessarily a person? Agentic commerce envisages AI systems moving beyond recommending products to searching, choosing and potentially purchasing them within rules established by their users. The payment itself becomes increasingly invisible. “People don’t wake up in the morning and say I want to make a payment,” Eacott points out. “They want to buy a cup of coffee, or they want to book a holiday.” But making the payment invisible does not make the responsibilities surrounding it disappear. Authentication, mandates, spending limits, consent and liability all have to work when an agent sits between customer and merchant. NatWest has highlighted the

Keeping the faith: Technology is changing payments, but trust remains the constant

Correspondent banking is alive and kicking, it just looks and feels different Simon Eacott, NatWest can improve cashflow visibility, while payment status information assists reconciliation. Removing the need to collect card or bank credentials can also reduce data-security exposure. Eacott resists framing A2A as a strategic move against card networks, though, as it is often portrayed in Europe. “I don’t think it’s necessarily a battle between account-to-account payments and cards,” he says. “It’s about customer choice effectively, and that’s both for merchants and for the customers themselves.” Cards have had decades to develop protections, chargebacks and an almost universally understood experience. Open banking payments will need similarly compelling services if they are to become intuitive and routine. “It is going to be around use cases as opposed to necessarily the technology behind it,” says Eacott, “and that’s where the growth will come.” 26 DISCOVERSIBOS2026

suspicious activity. Customers shouldn’t have to choose between security and convenience. “The best controls are always going to be the ones the customers don’t notice,” Eacott says. That brings the conversation full circle. Correspondent banking is alive. Cards remain deeply embedded in commerce as A2A grows. Cash, Eacott believes, will remain necessary for some customers. Conventional bank money could increasingly coexist with tokenised deposits, while AI agents may join humans in initiating payments. The future is therefore not a neat succession of old rails being removed as new ones arrive. It is a hybrid system in which interoperability matters more. That is why Eacott’s insistence on ‘stress testing’ the latest technology feels particularly apt at Sibos. Banks must think about what comes next while millions of transactions still have to move safely, securely and reliably today.

emerging concept of ‘Know Your Agent’, extending authentication into a world where machines transact with other systems. Eacott expects adoption to be slower than some of the excitement surrounding agentic commerce suggests. “If AI agents are going to be moving money, they need the same rules, they need the same boundaries, they need the same accountability,” he says. “You cannot automate away trust.” AI is also, of course, available to attackers, with deepfakes and AI-generated scams, and banks cannot solve that problem alone. “An awful lot of fraud does not start in the banking system,” says Eacott. “It’s the banking system which then facilitates the movement of money, but the fraud itself is taking place on other platforms.” That means bringing telecoms providers, technology businesses and social-media platforms into the response, sharing intelligence faster and using AI itself to identify

“Nobody rings you up and thanks you for a payment going well,” he jokes. Behind that line sits the central tension facing banks. Innovation can be rapid; infrastructure cannot afford to be reckless. Corporate customers want real-time visibility, better data and lower friction, but also expect resilience and certainty. Consumers may embrace AI-assisted commerce, but still expect somebody to take responsibility when it goes wrong. Perhaps the most important lesson of the past decade is not that payments innovation failed to fulfil its promise, but that financial networks evolve differently from consumer technology. The old infrastructure did not vanish. It connected to the new. And as tokenised money, open banking, instant payments and AI converge, Eacott believes the industry’s job remains remarkably consistent: focus less on the technology itself and more on what it enables. “Everything starts with trust,” he says. And for all the changes likely to dominate the agenda in Miami, that may be the underlying payments rail that matters most. FFNEWS.COM

DISCOV

S6 O SM IIAB 2 0 2 I M

Ahead of the numbers: AI is moving from assisting operations teams to taking on the exceptions themselves

RECONCILED TO THE TRUTH

POST-TRADE

As settlement cycles shrink and financial institutions wrestle legacy infrastructure, Smartstream believes only one type of intelligence can handle the pressure Four, three, two, one, zero! For decades now, the direction of travel in capital markets has been declining… in a good way. Settlement cycles have compressed from T+4 to T+1 and are heading towards all-or-none, atomic settlement (aka T+0). It improves liquidity, makes customers happy, but narrows the 28 DISCOVERSIBOS2026

window to identify discrepancies, investigate exceptions and put them right. And post-trade bank infrastructure, built for an era when information arrived in bulk overnight rather than continuously throughout the day, is struggling to catch up with this accelerated timeframe. “We’re living in a transitional world,” says Thomas Steinborn, Chief Product and Technology Officer at Smartstream. “The problem is not the straight-through processing. We solved that long ago.” Rather, he says, it’s exception management. A transaction that matches automatically requires little human attention. An exception can send an operations specialist searching across market data, internal dashboards, middle-office systems and counterparty communications for an

Thomas Steinborn, Chief Product & Technology Officer at Smartstream

explanation. Multiply that work across the transaction volumes handled by major financial institutions, and speed and accuracy are on a collision course. The US has already moved to T+1. The UK and EU are preparing for their own transitions. Processes that once waited until the following morning, now must happen within hours. Smartstream’s research among buy-side firms suggests much of the industry is a long way off achieving that, though. More than 70 per cent of respondents told it they still relied primarily on end-of-day reconciliation, while 53 per cent identified timing differences and data mismatches as the principal cause of reconciliation breaks. Almost seven in 10 described their preparations for T+1 as only partially adapted and still evolving.

FFNEWS.COM

Faster markets require faster controls. Yet the infrastructure feeding those controls does not automatically become real-time because the settlement deadline changes. “I’ve been in this business for the last 25 years – I can say it will never be resolved,” says Steinborn. Which is why Smartstream favours a different solution: agentic AI.

From copilots to colleagues The first wave of AI inside financial institutions was largely focussed on copilots that summarised information, answered questions or helped employees navigate data. Agentic AI represents a more consequential step because it can take actions and progress a task autonomously within defined approval rules. Smartstream has been developing that concept specifically for financial operations. Its Smart Agents are designed to investigate exceptions and undertake elements of work traditionally performed by operations teams: retrieving information, determining why transactions have failed to match, monitoring service levels, triaging issues and escalating cases requiring human judgement. “There will always be a new message format. There will always be a new rail. There will always be a new data feed. In other words, complexity regenerates, and AI offers one way of managing it,” says Steinborn. He’s not talking about one autonomous system replacing an entire department. “I need a team of agents,” he says. One might investigate a break, another manage triage, another monitor deadlines, while others handle escalation or approval. Humans remain involved where judgement and accountability matter. Smartstream says agents recruited by a Tier 1 financial institution have already proved their worth. Investigation time for an exception fell from around 14 minutes to 30 seconds in its trials, while other institutions projected that between 50 per cent and 70 per cent of investigation activity can potentially be automated during the first year.

The arrival of ‘industrial AI’ Steinborn says Smartstream was not initially being pushed particularly hard by clients to introduce AI. Interest existed, but budgets, governance structures and organisational readiness often did not. That picture is changing. Steinborn estimates that while around half have reached a stage where copilots should be deployable, roughly 10 per cent of Smartstream’s customer base already has agentic AI in production: “What is FFNEWS.COM

changing this year is really people moving to industrial AI and deploying agents.” That’s putting AI inside processes on which institutions actually depend, which raises the stakes. In financial services, an AI system producing a poor summary is one thing. An AI system operating inside transaction processing or post-trade controls is another. Governance, explainability, auditability and human oversight become part of the architecture. The challenge is not simply giving an agent sufficient autonomy to be useful, but constraining that autonomy sufficiently for a regulated institution to trust it.

The problem of data quality drift There is another prerequisite for that trust: data. For all the attention devoted to sophisticated models, AI remains dependent on the information it receives. Yet Steinborn argues that institutions frequently misunderstand the nature of the data problem. “The problem is not data quality,” he says. “The problem is data quality drift.” A data feed can be clean when a system enters production, a reconciliation process can achieve excellent matching rates, and then something changes – a field is altered, an upstream platform is upgraded, a counterparty changes its message, or a new feed appears, and what once worked almost perfectly begins to deteriorate. Operations teams compensate, manual intervention increases and straight-through processing declines.

THE CORE OF OUR PRODUCTS WILL BE RETHOUGHT WITH AI IN THE MIDDLE Smartstream sees AI taking on another role here: not merely doing the operational work but observing the infrastructure performing it. Rather than waiting for deteriorating data to generate a mountain of exceptions, AI systems can identify changing behaviour and alert teams before operational performance is compromised. Reconciliation then starts to evolve from a retrospective process – discovering what went wrong – towards something closer to an always-on control layer.

A new order Smartstream is already pushing its technology in that direction through Air, its AI-based reconciliation and data automation platform.

Recent deployments illustrate the potential. The National Bank of Greece’s Cyprus operation consolidated four reconciliation environments covering instant, cash, SEPA and nostro processes onto Air. Papua New Guinea’s Comrade Trustee Services also adopted the technology, removing the need for manual data collection, spreadsheet preprocessing and complex file handling. Smartstream says a process that might previously have taken eight hours was reduced to less than five minutes. But automating existing processes is only part of the ambition. Steinborn ultimately envisages something more fundamental: reimagining the architecture of Smartstream’s products by building around AI itself. “The core of our products will be rethought with AI in the middle,” he says, rather than ‘adding AI as an afterthought to the core architectures’. That shift may prove more important than any individual agent or copilot. Financial institutions have spent years attaching new technologies to infrastructure designed for another era. APIs were connected to batch systems. Cloud services were layered onto legacy applications. AI copilots are now being placed alongside workflows built long before large language models existed. An AI-native approach asks a different question: if the technology were being designed today, knowing what AI can now do, would the workflow look like this at all? For Smartstream, the answer increasingly appears to be no. The future it describes is one in which financial data is understood as it enters an organisation, reconciliation happens continuously, deterioration is detected before it becomes systemic, and intelligent agents investigate the exceptions that remain – escalating to people when human judgement matters. There will always be exceptions. There will always be new systems, message formats, rails and data feeds. Indeed, if Steinborn is right, interoperability may never finally be ‘solved’. But that may be precisely the point. For decades, Smartstream’s technology has been concerned with identifying differences between competing versions of financial reality and helping institutions resolve them. Now the company is confronting a rather larger reconciliation of its own: bringing decades of financial infrastructure into line with a market that is becoming faster, more automated and increasingly intelligent. The back office has spent years catching up with what happened yesterday; Smartstream is helping it operate in the now. DISCOVERSIBOS2026 29

After decades of disruption, the payments industry has accepted that change is the new normal. But dynamic interoperability can help absorb the shifts, says Bottomline Payments modernisation has created a curious problem. Banks have more ways to connect, more data to work with and more technology at their disposal than ever – but also a growing number of rails, systems, standards and services to manage. The Payments Intelligence_Gap 2026 Report from Bottomline suggests the industry remains caught between ambition and execution. Some 63 per cent of financial institutions surveyed are still only early-to-midway through digital transformation, while 40 per cent identify legacy infrastructure as the biggest barrier to real-time payments. Just 14 per cent consider themselves advanced in AI integration, despite 67 per cent planning to prioritise it over the coming year. Bottomline’s answer to this semi-paralysis is Global Pay Connect, its new software-as-a-service (SaaS) platform bringing connectivity, financial messaging, orchestration, compliance and fraud controls into a common environment.

The principle behind it is that connectivity itself is no longer the hard part. The challenge is creating an architecture that can absorb continuous change without creating another generation of silos. We spoke to Edward Ireland, Product Director, Financial Messaging, and Natasha Lapierre, Head of Financial Messaging Product Strategy and Innovation at Bottomline, about connecting old and new payment worlds, getting value from ISO 20022 and preparing AI for the point at which banks are prepared to trust it. DISCOVER SIBOS MAGAZINE Bottomline’s latest research gathered from more than 300 global payment professionals suggests 40 per cent of institutions still see legacy infrastructure as the biggest barrier to real-time payments. Why is the underlying problem proving so persistent – and how has the connectivity challenge changed? EDWARD IRELAND With our first customers, it really was all about connecting them into schemes and networks. Sending and receiving information was the payments business. The payments landscape and the requirements financial institutions now have are infinitely more complex than the connectivity itself. It’s all the other services around the payment rail that need to come into the picture: pre-validation, tracking, confirmation and overlay services, as well as fraud detection and sanctions screening. The ask isn’t just connectivity anymore; it’s a full-service offering. The legacy

No more

Edward Ireland, Product Director, Financial Messaging at Bottomline

Natasha Lapierre, Head of Financial Messaging Product Strategy and Innovation at Bottomline

NEW ARCHITECTURES

‘business as usual’ 30 DISCOVERSIBOS2026

problem comes from the way these environments have been developed. Payment environments were built in silos and, when a new payment environment came along, another silo was added. Institutions now have multiple payment rails to support simultaneously. That becomes a particular issue with instant payments. Historically, banks had more time within the payment workflow to apply controls. With an instant payment, anything you need to do around that transaction increasingly has to happen before you make it. Institutions need more controls, access to more databases, reporting and connections into new schemes. The important thing is that, as they implement all of this, they don’t create additional silos. Standards such as ISO 20022 give us common denominators between schemes. Instead of developing each rail independently for each market, institutions can bring them into common processes and environments. DSM That makes ISO 20022 more than a migration project. The major Swift coexistence milestone has passed, but the industry is still working through issues such as structured address data. Where does the real value come from now – and why can apparently straightforward industry-wide changes still prove so difficult? EI ISO 20022 still has a long way to go. We had the big cutover, but what we’re seeing now is much more about the implementation and use of ISO. Structured addresses are a good example of how difficult change at scale can be. It’s something the whole market has wanted to do; it has been flagged for years, and everybody is

FFNEWS.COM

DISCOV

working on it. But the more you dig into it, the more challenges you find. So much so that deadlines for structured addresses have been delayed from November 2026 and extended to a still undetermined date. Nobody is arguing that structured addresses are a bad idea – there are multiple benefits and they can drive efficiency – but getting an entire market there shows how difficult coordinated change can be. The upside is that institutions are also starting to appreciate how much commonality exists between different schemes. ISO 20022 has demonstrated the similarities in how information is identified and used across different markets. If you can bring that into a single platform or common teams, you can leverage expertise rather than maintaining expertise within every individual silo. There are economies of scale, but also the opportunity to take best practice and learning from one market into another. And the environment won’t stand still. New payment rails are being added all the time, particularly as we look towards tokenised deposits, stablecoins and other emerging ways of making payments. The old schemes don’t die: the new schemes get added while the old schemes remain.

Multiple networks, fintech partners and data providers are no longer a temporary state. They’re becoming the permanent shape of the ecosystem Edward Ireland, Bottomline DSM Bottomline’s research finds only 14 per cent of institutions consider themselves advanced in AI integration, yet 67 per cent intend to prioritise AI over the next 12 months. Where can AI create genuine value now and what has to happen before banks will trust AI to make payment decisions itself? NATASHA LAPIERRE There is a lot of potential, but the industry is at a particular stage of readiness. AI is exceptional at processing huge amounts of information in record time, identifying patterns that would

be difficult for a human to spot and surfacing ‘unknown unknowns’. So the potential range of use cases is enormous: anomaly detection, operational monitoring, intelligent orchestration, automated alerting and many more. But the reality today is that most banks aren’t ready to let AI be fully automated and make actual payment decisions. That is largely because they don’t fully trust the data they are feeding it in the first place. The use cases that are winning today are therefore those where AI accelerates a human decision rather than replaces it. That might mean improving the accuracy of sanctions or fraud alerts, or suggesting how a payment field should be enriched or adjusted to meet a particular market infrastructure or rulebook. But a human makes the final call. There are two important principles there. The first is keeping the human in the loop. The second is embedded explainability. If AI tells me ‘87 per cent’, what do I do with that? It needs to tell me why it is 87 per cent and where the underlying data came from, so what it provides is directly actionable. The other battle is data fragmentation. Payment data is siloed between schemes and systems and, in some cases, isn’t fed back into areas such as the fraud engine. There is fragmentation between banks, too. AI is ready. We need to feed it the right things. DSM With traditional rails, instant payments,

fintech partners, third-party data and digital assets all coexisting, is ‘future-proofing’ now less about picking the right technology and more about avoiding irreversible choices? NL Interoperability is no longer something that is nice to have or something to think about in the future. It’s foundational, and it’s happening today. The industry is coming to terms with the reality that multiple networks, fintech partners and data providers are no longer a temporary state. They’re becoming the permanent shape of the ecosystem and, if current trends continue, that is only going to increase.

ER S I B OS M

Banks often ask I A M I 202 ‘Which network do I 6 commit to?’, ‘Which partner do I integrate with?’, ‘What happens if I build an integration and it turns out to be the wrong one?’. Financial institutions need to transcend that question completely. The institutions that get this right won’t necessarily be those that build as many integrations as possible, or those that place the right bets – particularly when it comes to digital asset networks. They’ll be the ones that don’t have to make an irreversible choice because their infrastructure is natively interoperable from the start. APIs are fundamental to that because banks increasingly need not just technical agility but data agility. The story is no longer about connecting to a payment scheme; it’s about connecting to a payment ecosystem, including network data and third-party intelligence.

The institutions that get this right … will be the ones that don’t have to make an irreversible choice because their infrastructure is natively interoperable from the start Natasha Lapierre, Bottomline There was perhaps an impression around some of the big ISO 20022 deadlines that once you reached them you wouldn’t have to change anymore. That’s no longer reality. Change is becoming business as usual. Historically, launching a capability or integrating a new source of data could mean another hard-coded integration project. With an API-enabled platform, you flip that premise: build the connection once and each new capability can increasingly become a configuration choice rather than another new build. That’s what future-proofed architecture looks like. It isn’t reaching a point where you never have to change. It’s being built for continuous change.

Charting a course: Interoperability can help banks navigate a changing payments landscape

FFNEWS.COM

DISCOVERSIBOS2026 31

Connect with Irish innovation Enterprise Ireland supports and funds the most ambitious Irish companies, helping them bring groundbreaking Irish innovation to a global audience. With 42 offices across the globe, our industry advisors are here to connect you with the power of Irish innovation. Discover how we can help your business at enterprise-ireland.com

Helping Irish business go global

DISCOV

ER S I B OS M I A M I 202

SETTING LIQUIDITY FREE TREASURY

In helping to release treasury teams from the drudgery of manual processes, J.P. Morgan Payments is also liberating the cash that can be crucial to growth “Most clients don’t call us because they have a cash problem. They’re calling us because they have a complexity problem,” says Mali Bartlett, Managing Director for Liquidity & Account Solutions at J.P. Morgan Payments. And it doesn’t get much more complex than an airline company caught in the middle of an oil crisis. Earlier this year, Peter Donkersloot, CFO of Panama-based Copa Airlines, was watching the unfolding situation in the Middle East, his company’s spiking fuel bill, and wondering what action he should take to protect liquidity and working capital. So he picked up the phone to his bank. FFNEWS.COM

He wasn’t sure how his call would be received, but he was pleasantly surprised. “My role as CFO today is balancing … that short-term issue with the longterm priorities of the company and making sure that we come out of this situation on top,” says Donkersloot. “I pick up the phone and I get two, three, four people that are willing to talk to me, are willing to connect me to the experts within J.P. Morgan… And that’s very good for us because even though we’re a large consumer of oil, we don’t have a team of people tracking what’s going on. “Our relationship with the bank right now has to be much more than a transactional relationship, and I think J.P. Morgan excels in that.”

Mali Bartlett, Managing Director for Liquidity & Account Solutions at J.P. Morgan Payments

The insights the bank shared helped Donkersloot navigate the price shock, keep 119 planes in the air and press ahead with the company’s investment plans. Three months later, it agreed to buy 60, 737 Boeing Max Jets, Copa’s CEO citing strong liquidity and a conservative balance sheet as having provided the necessary flexibility to w eather the storm. The wraparound service that J.P. Morgan Payments offered Copa in a crisis is evidence of a reassuringly human touch, but the underlying insights were the result of significant advances in technology driven by rapidly changing transaction models that are fundamentally reshaping the treasury function. DISCOVERSIBOS2026 33

DISCOV

S6 O SM IIAB 2 0 2 I M

“A few years ago, treasury teams had more time to gather information, assess their positions and make decisions,” says Bartlett. “But today payments are moving much faster, businesses are operating around the clock, and expectations around real-time information have dramatically increased. “As a result, it’s no longer good enough to just ask how much cash you have. Treasurers need to know where it is, what currency it’s in, whether or not it’s accessible, and how quickly it can be mobilised when the business needs it. “What we’re hearing from many clients is that liquidity management needs to become more dynamic. Because it’s less about managing deposits and more about managing the movement of money across multiple markets and currencies, in response to that speed and those changing expectations.” J.P. Morgan Payments combines corporate treasury, trade, merchant services, and cross-border capabilities into a global engine, processing nearly $12trillion a day. It’s supremely good at transaction management, and providing the tools to treasury teams to accelerate and monitor that flow. But now, says Bartlett, it’s ‘moving towards a world where liquidity structures can react to information in real time and automatically position that cash where it’s needed and when it’s needed’. Simplifying the CFO’s job is a top priority for Bartlett as the person responsible for the design and creation of products that help treasury clients gain visibility and maintain the control they need. It means they ‘spend less time assembling the picture and more time acting on it’, she says.

Unlocking liquidity in time As companies grow, expand across borders, maybe add new legal entities, their biggest problem is not necessarily a lack of liquidity, but rather that cash is trapped where it can’t be accessed quickly enough – if at all. A business can look cash rich on paper and struggle to settle the bills. “The major challenge is about the trapping and fragmentation of the cash,” says Bartlett. “One of the misconceptions that we deal with is that a large cash balance can automatically be moved, and thus that the customer has plenty of

34 DISCOVERSIBOS2026

liquidity at hand. What we find is that the cash may be sitting in the wrong legal entity. It could be in a market with regulatory restrictions, or even in a currency that is difficult or costly to mobilise very quickly. “Even in a restricted market, cash doesn’t always have to be inaccessible. We’re helping clients think about structures that can improve visibility, increase the fungibility of it, and where possible automate the movement of liquidity across their organisation, and, combine liquidity management alongside FX solutions.” Earlier this year, another airline, the European low-cost carrier Wizz Air, became the first client to implement J.P. Morgan Payments’ virtual netting solution, which streamlines settlement of intercompany multicurrency invoices to reduce foreign exchange (FX) costs and manage risk exposure across a global operation. As an alternative to traditional netting software, it uses virtual accounts to settle inter-company invoices, using the in-house bank’s own FX rates to translate each separate invoice into the correct settlement currency and amount. The solution builds on J.P. Morgan Payments’ virtual account management and integrated FX

The opportunity is in moving towards a world where liquidity structures can react to information in real time and automatically position that cash where it’s needed and when it’s needed

solutions and, for international operators like Wizz Air, it could save millions of dollars a year in currency conversion and transaction fees. Bojtos Gábor, Head of Group Treasury at Wizz Air, was delighted to ‘stop paying banks to move their own money’ and said the solution had helped deliver ‘hard savings and radical simplification that fully aligns with our ultra-low-cost carrier principles’. Every improvement that J.P. Morgan Payments helps deliver is predicated on identifying where a client’s financial data sits and introducing a level of automation to manage it. Once that exercise is complete, the treasury function is liberated – just like the cash. “It really is about coming in each morning to see where their liquidity is concentrated, where their funding needs are emerging, and what actions they need to be taking,” says Bartlett. “That could be using solutions that we’ve developed, like programmable liquidity, an event-triggered sweep, where the movement is driven by the information and the guardrails, not just by the calendar and time.” So what does she believe CFOs in charge of a global ledger should be focussing on right now? “I think there are three questions. How do I get greater visibility into my global position? What automations can I introduce to remove the manual processing my team is undertaking today? And how can emerging technologies help my team make better and faster decisions? Ultimately that’s where the industry is heading, which will then allow treasury to become more connected, more intelligent and more strategic in the future.”

Mali Bartlett, J.P. Morgan Payments

FFNEWS.COM

Payment Infrastructure for Innov ators

Helping fintechs and payment businesses launch faster with issuing, digital banking, and compliance solutions.

Issuing Services

Digital Banking

Digital Banking Direct

Payments Compliance

Scalable card issuing and BIN sponsorship

Virtual business accounts and payouts

GBP & EUR accounts with real-time payments

AML and regulatory support

Ready to Scale?

Moorwand helps fintechs move faster with confidence. Contact Us > sales@moorwand.com Moorwand helps fintechs move faster with confidence.