The European Central Bank began a two-year preparation phase for a digital euro in November 2023. The Bank of England launched a digital pound consultation in February 2023. China’s e-CNY has already been tested in 26 cities with over 260 million wallets created. These are not fintech experiments. They are central bank initiatives that will require every commercial bank in their jurisdictions to integrate with entirely new payment infrastructure. The technology decisions banks make in the next three to five years will determine whether they participate in this future or are sidelined by it. The banking infrastructure market supporting these transitions reached $18.6 billion in 2024, according to Global Market Insights, growing toward $73.7 billion by 2034.
Central Bank Digital Currencies and Their Infrastructure Demands
A CBDC is a digital form of a country’s currency issued directly by the central bank. Unlike commercial bank deposits (which are claims on a private bank) or crypto assets (which have no central issuer), a CBDC is a direct liability of the central bank, as safe as physical cash but existing only in digital form.
For commercial banks, CBDCs create both a technology challenge and a competitive threat. The technology challenge is integration: banks will need to process CBDC transactions alongside existing deposit and payment systems. Their core banking platforms must be able to hold, transfer, and account for CBDC balances. Legacy mainframes built in the 1980s were not designed for this.
The Boston Consulting Group projects fintech revenues will reach $1.5 trillion by 2030, with embedded finance and digital lending accounting for the largest share of projected growth.
According to CB Insights’ 2024 fintech report, global fintech funding declined 40 percent between 2022 and 2024, pushing the sector toward consolidation and a sharper focus on profitability over growth at all costs.
The competitive threat is disintermediation. If consumers can hold digital currency directly with the central bank, they may reduce or eliminate their deposits at commercial banks. Central banks are designing CBDCs to mitigate this (holding limits, no interest payments), but the structural possibility of a direct central bank-to-citizen financial relationship changes the role of commercial banks in the monetary system.
Banks that run on modern, API-native platforms will integrate CBDC rails faster than those on legacy systems. This is another reason the BaaS market’s cloud deployment share has reached 67%, per Global Market Insights: cloud-native platforms are inherently more adaptable to new payment types.
AI in Banking Operations
Artificial intelligence is moving from a supporting tool to a core component of banking operations. Three applications are already deployed at scale.
Fraud detection uses machine learning models that analyse hundreds of transaction variables in real time: location, amount, merchant category, time of day, device fingerprint, and behavioural patterns. These models catch sophisticated fraud that rule-based systems miss and produce fewer false positives, which means fewer legitimate transactions are incorrectly blocked. Banks globally process over 2 billion API calls daily, handling $676 billion in transaction value, per Coinlaw. AI-powered fraud screening runs on every one of those transactions.
Credit decisioning increasingly uses AI models trained on transaction data rather than relying solely on credit bureau scores. These models assess cash flow patterns, spending behaviour, and income stability to produce more nuanced lending decisions. For thin-file borrowers (those with limited credit history), AI-based models are significantly more accurate than traditional scorecards.
Customer service automation has progressed from simple chatbots to AI assistants that can handle complex queries: explaining transaction charges, processing disputes, adjusting payment schedules, and providing personalised financial guidance. Banks that deploy these systems report handling 70% to 80% of customer enquiries without human intervention.
Programmable Money and Smart Contracts
The concept of programmable money, currency that can execute rules automatically, is moving from blockchain experimentation to mainstream banking discussion. A programmable payment could automatically split rent among roommates on the first of each month, release escrow funds when a delivery confirmation is received, or adjust insurance premium payments based on usage data.
Tokenised deposits, which represent commercial bank deposits as digital tokens on a shared ledger, are being piloted by JPMorgan (Onyx), Citi (Citi Token Services), and HSBC (Orion). These tokens can settle securities transactions instantly, eliminate the multi-day clearing processes that currently sit between trade execution and settlement, and enable 24/7 operation of financial markets that currently close on weekends and holidays.
The infrastructure required to support programmable money does not exist in legacy banking systems. It requires real-time processing, smart contract execution capabilities, and interoperability with both traditional payment networks and new digital asset rails. Banks on next-generation platforms are better positioned to participate.
Embedded Finance at Scale
Embedded finance, banking services distributed through non-bank platforms, will grow as the technology infrastructure matures. When a ride-hailing app offers driver loans, an e-commerce platform provides buy-now-pay-later, or an accounting software package offers business credit, a bank or BaaS provider supplies the financial infrastructure behind the scenes.
Platform-based BaaS models account for 69% of the global market, per Global Market Insights. The neobanking market reached $210.16 billion in 2025, per Fortune Business Insights, growing at 49.30% annually. Both numbers reflect the ongoing disaggregation of banking: financial services are separating from financial institutions and being distributed through the platforms where customers already spend their time.
The cross-border dimension amplifies this trend. The global cross-border payments market reached $371.59 billion in 2025, per Fortune Business Insights. As embedded finance goes global, the platforms distributing financial products need infrastructure that works across currencies, regulatory regimes, and payment networks.
The Convergence
Future banking systems will not be defined by any single technology. They will be defined by the convergence of all of them: cloud-native core platforms processing transactions in real time, AI models making decisions on every transaction, CBDC rails providing new payment options, programmable money automating complex financial logic, and embedded finance distributing banking services through every digital platform customers use.
The institutions that are already preparing include both neobanks built on modern stacks and traditional banks that began their modernisation programmes early enough to have modern infrastructure in place. The ones that waited are now making technology decisions under time pressure, with each passing quarter reducing their options and increasing their costs.
The banks that build for this convergence are investing in modular, API-first architectures that can integrate new capabilities as they emerge. The banks that treat each technology trend as a separate initiative, managed by a separate team with a separate budget, will find themselves unable to connect the pieces when the convergence arrives. It is arriving faster than most bank technology roadmaps assume.
