Technology Rewiring

Agentic AI is moving rapidly from concept to deployment, but its trajectory across financial services is far from uniform. Institutions are experimenting with autonomous agents across markets, functions and business models, while adoption, ambition and readiness vary significantly.
This opening conversation takes stock of the emerging global landscape: where is agentic AI gaining real traction, which parts of finance are moving fastest, and what are institutions learning as they move from experimentation towards deployment? Most importantly, as the technology develops at different speeds and in different ways, where could the next wave of value emerge?

As AI reshapes commerce, it is also reshaping financial crime. Deepfakes, synthetic identities and organised fraud networks are enabling attacks at unprecedented scale and speed. This session explores how security must shift from authenticating transactions to authenticating actors, human or machine. It also looks at why the organisations that thrive will be those that build the intelligence, visibility and response infrastructure needed to protect trust in an increasingly automated economy.

Every institution now has access to the same AI models, so durable advantage no longer comes from the technology itself—it comes from how deeply you rewire the business around it. This is the strategic half of a two-part conversation: where the growth, the margin, and the moat actually come from once agentic AI is table stakes.

The panel tackles the questions in the boardroom right now: What must you commit to in the next 12 months? What does a rewired financial institution look like in 2027—and what happens to those who wait? Where do you build proprietary advantage versus buy off the shelf?

Walk away knowing where agentic AI creates real competitive separation for your institution—and the strategic bets you need to place now to capture it before rivals do.

AI is changing how businesses buy, pay and move money. Systems can now interpret intent and orchestrate invoice approval, payment execution and reconciliation as a single journey within defined controls. This shift is not about faster payments. It is a move from managing transactions to managing outcomes. Hear practical examples, a framework for the agentic enterprise, and what it takes in trust and infrastructure to scale agentic commerce responsibly.

Financial institutions aren’t short of intelligence. They are short of agents they will let act. Running AI agents in regulated finance requires bounded permissions, traceable actions and defensible outcomes. The challenge is no longer just what AI can do, but how institutions can govern what it does and prove it.

While Western payment giants race to crack Asia's fragmented markets, the region's homegrown fintechs have quietly gone the other way, expanding into Europe and the US. A moderated conversation with two cross-border payments leaders explores what mastering Asia's complexity has taught them about winning on the global stage.

AI agents will not only serve customers and execute payments. They will also be used to impersonate customers, automate fraud and probe financial infrastructure at unprecedented speed. Traditional controls built around human behaviour and manual intervention may no longer respond fast enough. This session examines how financial institutions can deploy autonomous defence, distinguish legitimate agents from malicious ones and contain threats without blocking genuine transactions.

The strategy is only as good as the operating model beneath it. This is the second half of the conversation—the same agentic rewiring, seen from the engine room, where COOs and CTOs turn ambition into lower cost and higher efficiency at scale.

The panel gets specific about execution: What has to change in your operating model, architecture, and workflows now? What does a production-grade, agent-run operation look like in 2027? How do you handle orchestration, guardrails, and human oversight without stalling between demo and deployment?

Walk away knowing what it actually takes to run agents in production—and where the efficiency and cost gains are real today versus still hype.

As agentic AI reshapes routine decisions, what happens when low risk and high frequency decisions are more easily be delegated to agent? The interface that defined digital banking for a decade is quietly fading, as autonomous agents transact, compare, and choose on the customer's behalf. When humans step back from the screen, the entire logic of customer engagement must be rebuilt.

Learn how the customer touchpoint transforms when agents become the primary user; what it takes to stay visible and preferred in an agent-mediated world; and how to prepare your channels, data, and brand for an interface-light future.

Advice has always been sold on presence: the advisor who knows you and is there when it matters. Technology scaled transactions, not presence. Agentic AI changes that. An advisor backed by AI agents can continuously serve hundreds of clients. This session demonstrates agentic AI in wealth management beyond CRM and lead generation, empowering advisors holistically, and explores what it takes to deploy it in production.

Agentic payments need more than smarter models to go mainstream. They need a new money stack of real-time settlement rails, agent-level policy controls and governance that earns trust at scale. What will it take to move from pilots to everyday commerce?

Financial services is entering a new phase where AI systems don't just assist, they act, orchestrating decisions across compliance, fraud detection, and customer engagement in real time. This session examines how the convergence of agentic AI with tokenization, stablecoins, and agentic commerce is reshaping payments infrastructure and core banking, and what FSI leaders need to consider as they move from generative AI pilots to autonomous systems operating at scale.

Three forces are colliding at once—autonomous agents, programmable stablecoins, and real-time rails—and together they could rewrite how money moves entirely. As AI agents begin to transact independently and stablecoins gain regulatory footing, the boundaries between payments, settlement, and currency are dissolving faster than most institutions can react. Learn how agentic payments and stablecoins intersect to enable machine-speed money movement; what infrastructure and compliance shifts this convergence demands; and where the earliest commercial opportunities are emerging.

Instant payments can reach millions of the unbanked—but only if the economics work for everyone building the rails. Around the world, national real-time payment systems are proving that speed and inclusion can coexist, yet many struggle to balance universal access with commercial viability. Getting that balance right is now a defining test of payments strategy.

You will learn: how to design instant payment schemes that serve the underserved without eroding sustainability; which governance and pricing models drive long-term adoption; and what lessons successful national roadmaps offer to markets still building theirs. Miss this session, and you risk designing a system that scales in usage but never in value.

The Agentic Bank opens Day 2 by zooming out, geopolitics, energy retooling, dark factories, autonomous transport, before zooming in on the reckoning inside finance. Autonomous banking cannot run on 1960s core systems, fiat rails, or batch payments, it demands infrastructure built for agents that act, not advise. This keynote strips away the interface to unpack what's underneath, architecture, governance, and risk controls. The key question it looks to answer: what actually separates genuine transformation from a rebranded chatbot?

By 2027, most financial institutions will have adopted the same AI tools—and discovered that tools alone change nothing. The real divide is opening between firms that bolt technology onto legacy structures and those bold enough to rebuild how they actually operate. Org charts, decision rights, and the boundary between human and machine are all up for grabs. Learn how leading institutions are redesigning operating models around AI and agents, what structural changes turn transformation into durable advantage, and where incumbents most often stall.

Banks are racing toward agentic infrastructure. Few can rebuild core banking, payments and risk management at once without the upgrade itself becoming the systemic risk. As autonomous agents move from pilots into production, institutions face a real choice: accelerate the rebuild, or slow down until controls and supervisors catch up. The panel pressure-tests both paths: what a credible migration sequence looks like, where legacy risk controls break under autonomous decision-making, and when deliberately slowing down is the smarter strategic call rather than a failure of nerve.

Imagine an economy where software doesn't just recommend decisions—it makes them, negotiates them, and settles them, all without a human in the loop. That future arrives when agentic AI meets programmable money. Autonomous agents that can transact, and digital money that can execute logic on its own, together unlock machine-speed commerce that today's financial rails were never built to handle. Learn how agentic AI and programmable money combine to reshape payments, settlement, and value exchange, what new risks emerge when agents transact autonomously, and how institutions should prepare for an economy that runs itself.

Money moves globally, but the rails beneath it don't speak the same language. Instant payment systems, stablecoins, card networks, correspondent banking, and CBDCs are all racing to define how value crosses borders—and no single winner is emerging. The future isn't one universal rail; it's many, and the real challenge is making them interoperate without friction, fragmentation, or lost trust. Learn how the competing rails of global money movement are converging and colliding, what interoperability actually requires across systems and jurisdictions, and where the next chapter of cross-border payments is really being written.

Fragmentation is now the operating condition for global banking. Geopolitical realignment, diverging regulation, shifting rate cycles and technological disruption are splitting a once-integrated system into competing blocs and rulebooks. For some banks these ruptures are existential; for others, they are the opportunity of a generation. The difference lies in who reads the shifts early and moves decisively. For bank leaders, the session pinpoints which ruptures are reshaping banking's risk and growth map, how leading institutions are turning geopolitical and regulatory headwinds into strategic tailwinds, and where the next sources of advantage are emerging in 2027.

For decades, financial innovation meant new products on old rails. That lens no longer explains where control is concentrating. Money 1.0 digitised value. Money 2.0 networked it. Money 3.0 makes it programmable — value that carries its own logic, constraints and compliance, settling toward atomic, with AI agents as first-class transactors. But is this a structural shift, or digitisation in new clothes? This opening session looks to separate architecture from hype and asks the question that matters: How far can programmable money along with tokenized financial products go?

Stablecoins, tokenised deposits, tokenised MMFs and CBDCs are all racing to become the default form of institutional money, though none of them do the same job equally well. Four advocates, one per instrument get two minutes each to pitch to our moderator, who presses each on the trade-offs. Then the audience votes: what would you actually hold for payroll, for treasury, for cross-border? Come find out what counts as money when the answer depends on the job.

As TradFi moves on-chain, can ownership of regulated digital assets remain anonymous? Programmed Governance offers a new model where trusted identity, asset eligibility and policy rules are checked and validated before value moves. As money becomes programmable, this session explores the next frontier: rewiring policy for an era of programmable value.

Banks have spent a decade buying products like Cash management, Trade Finance, Supply Chain Finance, Lending; and stitching them together afterwards. The cost isn't just efficiency; it caps growth: fragmented data, duplicated workflows, and change that moves at the pace of the slowest system. This session argues that banking's next competitive advantage is architectural, and shows what institutions building on one foundation do differently

Tokenization pilots keep succeeding technically and failing to scale, because liquidity fragments across a dozen new rails instead of moving as one. This session makes the case that fragmentation, not technology, is the real risk institutions face next, and shows what banks moving past the pilot stage are actually doing differently to solve it. 

RWA tokenisation has a surplus of vision and a shortage of honest post-mortems. This session is the latter. Which asset classes are seeing real traction, and which are stuck in pilot theatre? How much settlement has actually been compressed and what breaks in the transition? Tokenised collateral promises mobility the legacy system can't match, but only if the legal frameworks hold. The question for banks, financial institutions and asset managers: who are the new investors, and what do they need to see before they allocate? Practitioners who have shipped and carry the scars, separate the working from the wished-for.

As trade corridors shift and USD liquidity tightens, SMEs are moving into new markets faster than financial institutions can keep up. This session explores what's changing on the ground and how banks can support SME growth across emerging-market corridors and beyond.

Commerce no longer stops at borders, but much of the infrastructure behind it still does. As commerce becomes increasingly digital, cross-border and interconnected, we are seeing greater convergence across payment networks, technology and local ecosystems. What will it take to bridge these divides and build payments infrastructure that is secure, scalable and capable of supporting the next generation of commerce?

As digital assets become increasingly integrated into the financial system, trust is becoming a competitive advantage. What does it take to build digital-asset infrastructure that consumers, businesses and institutions can rely on?

Your next customer won't be human. AI agents are becoming first-class transactors — browsing, deciding and paying with no person in the loop. When software initiates the payment, the contest shifts from the rail to the moment of intent: who owns it — the wallet, the platform, the merchant, or the agent's operator? What does an agent need before it can spend: identity, mandate, limits, recourse? Why do stablecoins and programmable money fit machine-initiated commerce better than card rails built for humans? And when an agent pays the wrong party on the wrong instruction, who is liable? The checkout is being rebuilt for machines. This session is about who gets to own it.

A hotel concierge earns trust by arranging the journey, not owning the car. This session asks what banks and wallets can learn from that model. Through examples from super-apps, digital banks and behaviour-based rewards programmes, it explores how embedded mobility can drive revenue and loyalty. Participants will leave with a practical framework for testing the economics and safeguarding trust when partners deliver the service.

Tokenisation doesn't fail on vision. It fails on plumbing. Behind every live corridor sits an unglamorous stack that decides whether it scales. Which settlement layer — L1, L2 or private and at what trade-off? How do you achieve legal finality across jurisdictions that don't agree? Which interoperability standards actually matter, and who is driving them? This session decodes the replicable components behind the case studies — the decisions that separate a production system from an expensive proof-of-concept.

For decades, payment systems have been designed around human behaviour. AI agents play by different rules. Unlike humans, agents do not care about rewards, interfaces or brand loyalty. They optimise for speed, cost, and reliability.

As agents begin making payments on behalf of users and businesses, they may transact at a scale and frequency never seen before, paying for everything from AI services and digital content to machine-to-machine interactions. The result could be millions of low-value transactions that challenge the economics and design of today's payment infrastructure.

Are our payment rails ready for this future? As PayNow Gen2 considers capabilities relevant to agentic commerce, this session explores whether existing payment rails can adapt, what role programmable money could play, and how payment infrastructure must evolve when machines start choosing how to pay.

Every institution talks about the need for interoperability but few speak honestly about what it actually costs, in regulation, in pricing, and in the multilateral negotiation it takes to connect one country's payment system to another's. This panel brings together the messaging networks, reserve-pooling mechanisms, national schemes and multilateral hubs doing this work in practice, to work through the frictions: which regulatory frameworks still block genuine interoperability, how the economics of a cross-border link actually get split between the parties that build and operate it, and what a smaller or newer market needs in place before it can plug into a bigger one.

This panel aims to present a grounded view of where multilateral payment cooperation is actually working today, what it costs to get there and dive deeper into whether the real barrier to cross-border payment connectivity is technical, regulatory or commercial.

For CFOs and treasurers, programmable money is not a thesis, it's an allocation decision. Where does settlement risk really sit when value moves in real time, and who carries the exposure? How does tokenisation rewrite collateral management and intraday liquidity? What does the yield/risk trade-off look like across deposits and other forms of money, and what happens to deposit insurance and the lender of last resort when money is code? The question boards are starting to ask: whose balance sheet does each instrument sit on, yours, your bank's, or nobody's? And are balance sheet light banking models the safer answer? A working session for the people who answer with real money.

We built payment systems for humans—accounts, cards, and checkout flows designed around people clicking "buy." But AI agents don't shop that way. As autonomous systems begin to transact, negotiate, and settle at machine speed, they expose how much of our economic infrastructure simply wasn't built for them. Payments are just the surface; identity, trust, contracts, and settlement all need reinventing. Learn what economic infrastructure AI agents actually require beyond payment rails, how identity and trust must evolve when machines transact autonomously, and who is building the foundational layers of the agentic economy.

For years, payments competed on the rails—faster, cheaper, more reliable ways to move money from A to B. But as transactions become instant, ubiquitous, and nearly free, the rails themselves are commoditizing. The value is migrating elsewhere: to data, embedded experiences, working capital, fraud intelligence, and the software wrapped around every transaction. Moving money is table stakes; the real margin lives above it. Learn where payments value is genuinely being created as core processing commoditizes, how leading players monetize data, software, and embedded finance beyond the transaction, and what it takes to climb the value stack.

As AI in finance shifts from advising to acting, institutions face a speed–trust paradox. Using MAS' Safeguards for Agentic Finance at Runtime (SAFR) framework as a lens, this discussion explores why the data moat must evolve from volume to an operational foundation that carries trusted identity, mandate, disposition and audit data across agents, workflows and jurisdictions.

Financial crime doesn't operate in silos, yet most fraud detection tools do. Today's landscape is filled with point solutions: KYC platforms that stop at onboarding, AML engines focused on transaction monitoring, and network graph tools that excel technically. Fraud signals are everywhere, but data remains fragmented. AI Fraud Surveillance unifies financial, qualitative, and network signals into a single view, enabling fraud detection before it happens, not after.

In a world of synthetic everything, proof is becoming the scarcest commodity in finance. As AI-generated identities, deepfakes, and convincing digital deception become routine, institutions can no longer assume authenticity; they have to verify it. This session frames the problem and then works through the answer.

It opens with the shift underneath the shift: once trust can no longer be assumed, the burden moves from catching fakes after the fact to proving authenticity before a transaction happens at all. That reframes what needs to be provable rather than assumed, and who is positioned to build the systems that prove it. The discussion then gets specific about the different bets being placed on how to solve this: catching synthetic fraud at the point of verification, proving identity cryptographically without exposing the underlying data at all, and building the credential infrastructure that ties it together.

Attendees hear where each approach actually holds up against a motivated forger, and where the gaps between them still leave room for one to slip through.

AI's real blocker isn't the model, it's unstructured data. This live demo shows an agent that scans external market signals, matches them to an advisor's book, and turns a headline into a qualified proposal, so advisors arrive don't just walk in prepared, they proactively create the next opportunity. See Microsoft's Work IQ, Foundry IQ, Fabric IQ and Web IQ orchestrating as one platform to make this a reality.

While most payments conversations focus on consumer rails, the real bottleneck may be commercial: B2B and merchant payments across Asia, the Middle East and Africa remain fragmented and cash-heavy. This panel unpacks what it takes to build commercial payment corridors that actually work and where the friction still lies.

Fraudsters didn't wait for a governance framework. Autonomous agents are already probing payment rails, spinning up synthetic identities and running social-engineering scripts at machine speed, while many institutions still triage alerts in batches. With instant payments and agentic commerce scaling fast, the window between attack and response is now measured in milliseconds. The panel looks at where real-time controls are keeping pace and where they are structurally behind. It also covers how to tell a legitimate customer's agent from a malicious one, and what defending with agents looks like when regulators still expect a human in the loop.

The protocol race is over faster than anyone expected. Visa, Google, Mastercard, OpenAI, Stripe and Ant International have all shipped agent-initiated payment infrastructure, while countries such as Korea alone recorded over 100 million AI-to-AI payments in seven months. This discussion will be around where machine will transact first, at what granularity, and on whose stack.

That question does not have a global answer, because agentic commerce is gated by four preconditions that are unevenly distributed or agreed upon.

Identity: can an agent be issued a provable, scoped, revocable mandate.

Acceptance density: is there enough merchant coverage for an agent's choice to be meaningful.

Settlement granularity: can value move at machine speed and sub-cent size.

Customer trust and value: will people actually delegate? Does the result make everyday finance simpler, more transparent and more accessible, including for customers that conventional banking has overlooked?

The first three are being solved on the rails; the fourth is decided in the customer relationship. This panel puts both views on one stage: the platforms building agentic infrastructure, and a digital bank that rebuilt banking around customer needs. Together they map the next eighteen months by looking look at where AI creates practical customer value first and how institutions should balance automation with human judgment and accountability.

Payments are becoming faster and cheaper, but where will the industry make money next? This panel looks at where new revenue pools are emerging across the payments value chain — from data and embedded services to cross-border flows and new forms of money — and which business models can capture them from 2027 onwards.