Roland Schilling looks at whether AI can create truly frictionless cross-border payments

Cross-border payments are the backbone of global commerce. Every day, businesses pay overseas suppliers, banks transfer funds across jurisdictions, and individuals send money to family abroad. Despite decades of technology progress, cross-border payments remain expensive, slow, fragmented, and very labour-intensive.

AI is emerging as a powerful tool to address many challenges. Rather than replacing existing infrastructure, AI can act as an intelligent orchestration layer to optimise decisions, automate processes, and improve efficiency through the payment lifecycle.

MAKING COMPLIANCE MORE EFFICIENT

Compliance is one of the most complex aspects of cross-border payments. Financial institutions must comply with anti-money laundering regulations, sanctions screening, KYC, fraud prevention, and jurisdiction-specific regulations. Traditional rule-based systems often generate false triggers, which require manual investigation and delay legitimate payments. AI can offer a more intelligent approach by evaluating transactions in context rather than relying solely on static rules. Machine learning models can assess customer behaviour, transaction history, counterparties, geography, and payment purpose to build a dynamic risk profile.

OPTIMISING FOREIGN EXCHANGE

FX pricing is often the least transparent cost in international payments. Businesses frequently accept quotes without understanding if it is the best option available. Pricing is influenced by market volatility, available liquidity, and timing. AI can improve FX management by analysing historical trends, real-time market data, market depths, and transactional patterns to recommend optimal execution timing and strategies. Machine learning models can also forecast short-term currency movements, helping treasury teams to determine whether to execute immediately or wait for more favourable conditions, whether to execute an entire payment or split it into multiple orders. While AI cannot predict markets with certainty, it can identify probabilities and patterns that improve decisionmaking, reduce unnecessary costs, and help with more efficient hedging strategies.

SMARTER LIQUIDITY MANAGEMENT

Large multinational companies often maintain cash balances across multiple currencies and jurisdictions to ensure payments can be executed without delay. By prefunding accounts

in advance in multiple countries, working capital gets tied up for longer, which could be invested or deployed somewhere else. AI can help improve liquidity management by forecasting future payment volumes, seasonal demand, expected currency requirements and regional settlement obligations. This will enable the treasury teams to optimise funding requirements, reducing idle cash balances and minimising unnecessary borrowing. Not only can AI help reduce the need for extra cash, but it can also identify liquidity shortfalls before they become expensive. Delayed payments can not only cost money but also damage the organisation’s reputation.

INTELLIGENT PAYMENT ROUTING

Today’s global payments ecosystem consists of multiple payment rails, including traditional correspondent banking networks, real-time payments systems, card networks, mobile payments, digital wallets, and blockchain-based settlement solutions. Each method differs in cost, speed, reliability, and geographical coverage. AI can enable dynamic routing by evaluating multiple variables in real time, including transaction value, geographic destination, available liquidity, cost, and timing. Instead of using a single default route, an AI-powered payment orchestration platform can determine the most efficient path for each payment. Lower value payments may prioritise speed, where high value corporate transfers may focus on cost and settlement risk.

AUTOMATING RECONCILIATION

Reconciliation remains one of the most labour-intensive aspects of cross-border payments. Inconsistent or missing payment references and payments arriving from multiple systems leads to finance teams having to spend considerable time manually matching payments with invoices and investigating exceptions. Machine learning models recognise patterns across payment data, invoices, and accounting records to match transactions automatically, even when reference fields are incomplete or inconsistent. When exceptions occur, AI can identify the probable cause, recommend corrective actions, and prioritise unresolved cases based on business and cost impact. This will reduce manual effort while improving financial accuracy and accelerating reporting processes.

PREDICTIVE OPERATIONS AND RISK MANAGEMENT

One of AI’s greatest impacts is its ability to predict issues before they affect customers. By analysing historical payment performance and operational data, AI can identify patterns associated with payment failure and settlement delays. For example, if a particular payment rail is experiencing increased processing delays, AI can detect the trend early and recommend alternative routing of upcoming payments before any disruption occurs. Moving from reactive to proactive risk management will improve customer experience and reduce operational cost and reputational damage.

AI AS AN INTELLIGENT ORCHESTRATION LAYER

The most significant impact AI can make is to bring together multiple decision points into a single platform. Every cross-border payment involves multiple considerations: regulatory compliance, FX pricing and execution methodology, liquidity availability, payment routing, operational risk, customer preference, timing, and overall cost. All the above decisions are made across different teams. An AI-powered orchestration engine can evaluate these factors simultaneously and determine the optimal strategy for each payment. Rather than replacing banks or payment networks, AI coordinates them more effectively, resulting in fewer manual decisions, lower operational costs and reduced internal resource requirements.

WHAT DOES THE FUTURE HOLD FOR AI WITHIN THE CROSS-BORDER PAYMENT WORKFLOW

Despite the considerable potential, AI will not eliminate every source of friction in cross-border payments, as the greatest challenges are structural rather than technological. The future of cross-border payments is unlikely to be defined by a single technology or payment rail; instead, it will be shaped by intelligent orchestration across an increasingly diverse financial ecosystem. Different countries maintain different regulatory frameworks and capital controls, tax requirements, and licensing regimes. Settlement ultimately depends on financial institutions, payment infrastructure and central bank systems operating across multiple jurisdictions. AI can accelerate decision-making, but it will not be able to override legal or regulatory constraints. Truly frictionless crossborder payments may never be entirely achievable; however, AI has the potential to remove much of the operational friction that businesses experience today. By making international payments faster, smarter, and more transparent, AI is poised to play a vital role in the next generation of global financial services.