The Future of Banking: Open Finance and BaaS

The democratisation of finance is no longer a prediction. It is the architecture underneath the banking products customers already use. Open banking APIs, embedded finance rails, banking-as-a-service platforms, and real-time payment networks have quietly moved financial services from monopolistic intermediation to composable infrastructure. The banks that win the next decade will not be the ones with the biggest branch networks or the most patient risk committees. They will be the ones shipping the best software.

From monolithic bank to composable stack

For a century, a bank was a single vertically integrated stack: a licence, a ledger, a branch, a product catalogue, and a customer base. The last fifteen years have unbundled every layer of that stack. Core ledgers are now offered as-a-service by vendors like Thought Machine, Mambu, 10x Banking, and Tuum. Card issuing is an API call to Marqeta, Lithic, or Stripe Issuing. KYC, AML, sanctions screening, and fraud scoring are modular services with clear SLAs. Customer interfaces have decoupled from product manufacturing, so the bank that makes the product is often not the brand that sells it.

What this means operationally is that a FinTech in 2026 can go from concept to live product with a banking licence partner, a modern core, an issuing processor, and a compliance layer in months rather than years. Established banks that still run on mainframe cores face a structural choice: migrate the core, wrap the core with modern APIs, or cede category leadership to challengers that start with the modern stack.

Open banking has matured into open finance

The first wave of open banking was about payment initiation and account information, driven by PSD2 in the EU and the CMA9 mandate in the UK. That wave is now settled infrastructure. The next wave, open finance, extends the pattern to mortgages, pensions, investments, and insurance, and is being codified in the EU's Financial Data Access framework (FIDA) and in parallel regulatory motions in the UK, Australia, and Brazil.

For engineering teams, the practical implication is that data portability is becoming a default expectation. Customers will increasingly demand the ability to pull their financial history across providers and plug it into budgeting, credit, and advisory tools of their choosing. Banks that treat open finance as a compliance obligation will lose relevance. Banks that treat it as a product surface will find new ways to acquire and retain customers.

Embedded finance is the dominant distribution channel

The highest-growth segments of financial services are now embedded inside non-financial software. Merchant cash advances originate inside Shopify. Insurance is bundled into Tesla's vehicle experience. Payroll and tax financing are native features of Rippling and Gusto. Expense management cards issue from inside accounting platforms. The common pattern is that the customer never visits a bank website. The financial product lives where the customer already spends time.

The banks and FinTechs enabling this embedded distribution are the banking-as-a-service (BaaS) providers: Column, Unit, Treasury Prime, Solaris, ClearBank, and a short list of chartered banks that have built API-first sponsor programmes. Regulatory scrutiny of BaaS tightened considerably through 2024 following enforcement actions in the US, and the surviving providers are the ones with mature compliance, capital, and technology controls. For a platform evaluating BaaS today, the due diligence bar has shifted from commercial terms to regulatory resilience.

Real-time payments are reshaping the product catalogue

The US launched FedNow in July 2023. The UK has run Faster Payments for over a decade. The EU's SEPA Instant Credit Transfer mandate, which took full effect in January 2025, now requires every euro payment service provider to offer instant, 24/7 euro transfers at no surcharge over standard transfers. India runs UPI at a scale that dwarfs every other payment rail on earth.

For product teams, real-time settlement changes the maths of credit, liquidity, and fraud. Authorisation windows collapse. Fraud scoring must complete in sub-second budgets. Reconciliation and chargeback workflows that assumed batch settlement need to be re-engineered. For CFOs and treasurers, real-time payments change working capital management and the economics of short-duration credit. Banks that have not yet mapped their product catalogue against real-time rails are already behind.

Digital assets have found a regulated lane

The noisy speculation cycle of the late 2010s has given way to a quieter, more durable story. Stablecoins have become a meaningful cross-border settlement rail, with tokens like USDC and EURC processing trillions in annual volume. The EU's Markets in Crypto-Assets (MiCA) framework, with full application from December 2024, has set a clear regulatory perimeter for issuance, custody, and trading. Tokenised deposits and tokenised treasuries are live products at banks including JPMorgan, HSBC, and BNY. Asset tokenisation pilots have moved from consortia experiments to pilot-production transitions in funds management and trade finance.

What this means for banks is pragmatic. Digital asset capability is no longer optional for any institution serving cross-border corporates, asset managers, or high net worth clients. The build-or-partner question for custody, tokenisation, and stablecoin rails has moved up the roadmap.

AI is the fastest-moving layer of the stack

Generative AI went from boardroom novelty to production workload inside eighteen months. In banking software today, AI agents drive customer service case deflection, compliance drafting, fraud investigation, personalised financial guidance, and, in our client work, a 20 to 40 percent productivity lift inside engineering teams. The banks that will benefit most are not the ones with the biggest AI budgets but the ones with the cleanest data platforms and the sharpest governance.

The competitive dynamic is moving quickly. A challenger that ships a better AI-driven money coach, a better compliance review agent, or a better underwriting explainability tool wins market share that would have taken an incumbent years to contest through traditional channels.

The architectural decisions that will define the next five years

Looking across the banks and FinTechs Scrums.com has worked with over the last year, five engineering decisions consistently separate the institutions that are pulling ahead from those that are falling behind.

1. Core strategy. Banks that have committed to a modern core, either through migration or through a strangler-pattern wrap, can ship product changes in days. Banks that still own end-to-end on mainframe cores ship in quarters.
2. Data platform maturity. A unified event stream, a governed feature store, and a well-maintained customer 360 are the foundation for every AI, personalisation, and real-time product feature that follows. Skipping this layer is the single most expensive mistake a digital programme can make.
3. API surface. A bank that exposes its own products through a modern, well-documented API layer internally can compose new propositions rapidly. A bank that still integrates through point-to-point connections cannot.
4. Operational resilience. With DORA (EU Regulation 2022/2554) in force across the EU since January 2025 and equivalent expectations rising globally, operational resilience has shifted from back-office discipline to board-level KPI. Architecture choices, vendor selection, and exit testing all now touch this theme.
5. Talent model. In-house engineering capability remains the single biggest constraint on bank software ambition. The banks that are winning are the ones that have combined strong internal engineering leadership with specialist external partners for capacity, velocity, and domain depth.

What this means for product leaders right now

Three moves matter more than the rest.

1. Audit where your product roadmap depends on infrastructure your team does not own, and decide whether you are comfortable with that dependency for the next five years. If not, move.
2. Treat open finance, real-time payments, and AI as product opportunities, not compliance deadlines. The teams that ship first in each category will define the category.
3. Invest in the unglamorous layers: the data platform, the internal API surface, and the engineering discipline that keeps software safe at scale. Every flashy front-end feature eventually depends on them.

How Scrums.com helps banks and FinTechs build the next chapter

Scrums.com partners with banks, credit unions, and FinTech platforms to design and deliver the software that defines modern financial services. Our banking software development services span core modernisation, open banking and open finance APIs, embedded finance platforms, real-time payments integration, and AI feature delivery, all engineered for the security, resilience, and regulatory expectations of the sector.

If you are planning the next phase of your banking or FinTech platform, start a project with us.

FAQ

What is open finance and how does it differ from open banking?

Open banking is the first wave of customer-permissioned data sharing, focused on payment accounts and driven by regulations like PSD2 in the EU and CMA9 in the UK. Open finance extends the same pattern to mortgages, pensions, investments, and insurance, and is being codified in the EU's FIDA framework and equivalent regulations in the UK, Australia, and Brazil.

What is banking-as-a-service and why does it matter?

Banking-as-a-service (BaaS) is a model in which a licensed bank offers its products through APIs, allowing non-bank platforms to embed financial services into their own customer experience. It matters because most growth in financial services distribution now happens inside non-financial software, making BaaS the primary channel for modern product launches.

How are real-time payments changing banking software?

Real-time rails like FedNow, Faster Payments, SEPA Instant, and UPI collapse authorisation and settlement windows from hours or days to seconds. Fraud scoring, liquidity management, reconciliation, and chargeback workflows all need re-engineering. Product catalogues built for batch settlement lose relevance quickly.

Should a bank migrate its core platform or wrap it with modern APIs?

Both are valid. Migration offers the cleanest long-term outcome but is expensive, multi-year, and high-risk. Wrapping with a strangler pattern preserves the existing core while exposing modern APIs and enabling faster product iteration. The right answer depends on the bank's appetite for execution risk, the condition of the legacy core, and the competitive pressure in its market.

What role does AI play in the future of banking?

AI is now a production workload in banking, driving customer service case deflection, compliance drafting, fraud investigation, personalisation, and engineering productivity. The banks that will benefit most are those with clean data platforms and mature governance, not those with the biggest AI budgets.