The Future of Money: Better Outcomes, or Just More Complexity?

Digital assets, real-time payments and agentic AI all promise a financial system that is faster, richer in data and more flexible. For corporate treasurers and CFOs, however, the more important question is whether these innovations create clearer decisions, stronger control and better business outcomes. Phil Carmalt and Sid Gupta of Bank of America, and Sharon Toh of Swift, share their perspectives.

If strategic treasury is increasingly defined by its ability to improve decisions, then innovation has to be judged by a practical test: does it give the business better visibility, more reliable execution and greater confidence in uncertain conditions?

That test matters because treasury has never operated in a perfectly connected world. Money does not move through one clean channel; data does not arrive perfectly formed; and liquidity decisions are rarely made with full visibility across every bank, entity and market. Even in sophisticated organisations, treasury still depends on many small connections holding together – between systems, payment rails, operating models, regulatory requirements and what the business needs to achieve.

Against that backdrop, the current wave of financial innovation is both compelling and uncomfortable. Tokenised deposits, stablecoins, instant payment rails, digital ledgers and agentic AI point towards a system that could be faster, richer in data and more flexible. Yet a faster payment still needs certainty. A digital asset still needs interoperability. An AI-generated recommendation still needs accountability. Without those foundations, progress can begin to look less like simplification and more like an additional layer of complexity on top of an already fragmented landscape.

From concept to treasury use case

Stablecoins have already moved beyond the experimental stage in parts of the financial ecosystem, supporting liquidity for crypto trading, offering a proxy for US dollars in volatile economies, and giving consumers alternative ways to pay or be paid. “Stablecoins are here to stay because they have real-world use cases much broader than corporate and institutional payments,” says Phil Carmalt, Head of APAC Treasury Product, Global Payments Solutions, Bank of America.

For corporate treasury, the near-term applications are narrower but potentially meaningful. One of the clearest is 24/7 intragroup money movement, particularly where funds move between entities within a single bank. The broader treasury value of digital money increases when it can move between banks, jurisdictions and systems with the same confidence treasurers expect from established fiat bank payment infrastructure.

Bank of America is participating in several industry initiatives in this area, including Swift’s work on shared ledger. “It might be a world of many rails working together, addressing different use cases,” says Sid Gupta, Head of APAC FIG and Co-head of Global Non-bank Financial Institutions, Bank of America.

Bank of America is also building 24/7 treasury solutions without blockchain i.e. on the fiat rails, because the underlying client requirement – more flexible access to liquidity – is clear. For treasurers, that distinction is important: the value lies less in the rail itself than in the business problem it solves.

Interoperability becomes the real infrastructure

As more rails emerge, the infrastructure question shifts quickly from the assets themselves to how they interact. Fiat systems, tokenised deposits, central bank digital currency (CBDC) initiatives and distributed ledger technology (DLT) networks may coexist for some time, each working effectively within its own environment. The difficulty begins when value has to move across networks without creating new operational breaks.

_“The unlock is not the assets themselves – it’s the connective tissue around them,” says Sharon Toh, Head of ASEAN Region, Swift. New forms of value become useful at scale only when they connect into the systems corporates already use to manage liquidity, approvals, data, compliance and risk. Otherwise, innovation creates another silo: technically interesting in a narrow setting, but operationally burdensome once treasury has to reconcile, control and govern it.

Swift’s shared ledger project is designed around this gap. Toh is clear that Swift is not launching a token or replacing existing infrastructure; its role is orchestration. Three principles guide the project: interoperability over reinvention, orchestration rather than issuance, and a pragmatic focus on production-ready value. The first use case centres on tokenised deposits, with an initial group of banks due to go live in July. “The goal is that institutions can adopt digital assets without breaking how they already operate today,” Toh explains.

Real time needs a business reason

The same pragmatism applies to real-time payments. Faster settlement can be transformative when a payment is closely linked to a commercial process – e-commerce settlement, insurance claims, supplier payments that trigger the immediate release of goods. Many routine treasury flows are different. “Speed for the sake of speed, unfortunately, can give rise to increasing fraud and cyber security risks,” Carmalt warns.

Gupta adds a distinction that is often blurred: real time and 24/7 availability are not the same thing. A payment may settle instantly during operating hours, while true 24/7 treasury requires an organisation to support compliance, liquidity monitoring, approvals and exception handling outside traditional working patterns.

Swift’s new payments framework addresses this by building in upfront transparency on fees and FX, no deductions along the chain, and real-time visibility from initiation to final credit. “Payments is not just about being faster,” Toh says. “It is really about being more transparent and more predictable.” For treasury, that predictability may be as valuable as speed itself.

AI is useful before it becomes autonomous

AI is sometimes discussed as if treasury is waiting for a future point of arrival, when many applications are already embedded in day-to-day activity, from cash forecasting support and receivables reconciliation to fraud detection. Bank of America’s virtual assistant Erica has had measurable impact: Gupta estimates that without it, the bank would have needed around 11,000 additional people to support the same client base.

Agentic AI introduces a more complex set of questions. Carmalt sketches a possible future in which an AI agent produces a cash forecast, identifies excess deposits, obtains bank quotes and places a fixed deposit automatically. The technology may be capable of this before governance models are ready to support it. “AI cannot be held accountable for decisions,” Carmalt adds.

That distinction will shape adoption. Treasury’s role is to manage risk and protect the company’s interests, which means recommendations can be automated more readily than responsibility can be delegated. AI will take on more analysis, monitoring and preparation; human judgment remains as the key to decisions that affect uncertainties around funding, liquidity and risk management.

Fraud, fragmentation and the singleness of money

The same technologies that support efficiency can also expand the threat landscape. Faster payments reduce friction for legitimate users, but they also reduce the time available to detect and stop fraud. Deepfake video calls are among the most unsettling examples, combining technological sophistication with familiar pressure points: authority, urgency and the reluctance to challenge someone senior. Gupta advises staff to validate unusual requests through at least one additional source. “It is important to always take a step back,” he says.

Across markets, fragmentation remains the working reality. Domestic real-time payment rails exist widely but are not uniformly connected. As stablecoins and tokenised deposits proliferate, treasury teams may also need to think more carefully about the fungibility of money – whether value held in one form, on one rail, or with one issuer is truly equivalent to value held elsewhere.

Carmalt compares the current environment with the US banking era of the 1800s, when regional banks issued their own banknotes and users had to consider the creditworthiness of the issuer. “People tend to take for granted the singleness of money.” Issuer risk, redemption risk and settlement risk have not disappeared; they can return through a different instrument.

Innovation must earn its place in treasury

The opportunity is real. Digital assets could make liquidity more flexible. Real-time payments could improve commercial flows. AI could sharpen forecasting, strengthen controls and reduce manual work. Richer standards could make payments more transparent and easier to reconcile.

But the path from innovation to treasury value runs through interoperability, governance and trust. For CFOs and treasurers, the question is not whether the future of money will be more digital, faster or more intelligent. It is whether those capabilities help the organisation make better decisions, protect value and operate with greater confidence in a changing world. Without those foundations, treasury may simply inherit a faster, smarter and more fragmented version of the complexity it already manages today.

That is why innovation in treasury cannot be treated as a technology agenda alone. It becomes meaningful when it improves resilience, sharpens execution and gives the business more usable options. In that sense, the future of money will be better only if it is designed around the practical outcomes that treasury is now expected to deliver at scale.