Top 5 Challenges in Cross-Border Payments — and How Cognitive AI Can Solve Them

For fintech leaders, chief risk officers, and compliance professionals, cross-border payments often feel like a balancing act between speed, cost, and compliance pressure. While the sheer scale is staggering—global cross-border payment flows totaled about $194.6 trillion in 2024—so too are the complexities: new AML regulations, customer expectations for real-time payments, and ever-smarter financial crime networks.

Fast-growing fintechs and mid-sized banks in particular face a unique predicament—delivering speed and cost efficiency in cross-border corridors while maintaining robust financial crime detection—can feel like trying to steer a speedboat with cruise ship controls.

Here’s the good news: advanced technologies—particularly Cognitive AI—are helping institutions flip this script. But before we get there, let’s ground the conversation in reality.

The Cross-Border Payments Challenge in 2025

The cross-border payment landscape isn’t just growing. It’s evolving under pressure:

• • According to McKinsey’s 2023 Global Payments Report, cross-border payments revenue hit $250 billion, with fintechs capturing over 15% of the market.

• • Regulatory scrutiny has intensified, especially post-2024. The U.S. Treasury’s FinCEN flagged over $1 billion in suspicious cross-border activity linked to crypto rails and fintech intermediaries.

• • High-risk markets face additional headwinds. Mexico, for instance, saw sanctions against three financial institutions in early 2025 related to money laundering concerns tied to fentanyl trafficking. These sanctions highlight how cross-border payment channels can be exploited by criminal networks. In Nigeria, the Central Bank’s new AML automation guidelines released in May 2025 signal growing urgency for tech-enabled compliance.

Given that context, let’s break down the top five challenges fintechs and mid-tier banks grapple with—and where Cognitive AI is emerging as a strategic ally.

1. Fragmented Data Across Jurisdictions

The problem: Cross-border payments pull data from multiple systems: SWIFT messages, correspondent banking records, customer onboarding systems, transaction monitoring platforms. Yet, these data streams often exist in silos, making it difficult to build a unified risk profile.

The compliance risk: Regulators now expect holistic customer and transaction risk assessments. The EU’s AMLA Authority, coming into effect in 2028, explicitly emphasizes integrated monitoring across business lines and jurisdictions.

Where Cognitive AI helps: Unlike rule-based systems that flag only pre-coded patterns, Cognitive AI systems can ingest and correlate fragmented data—structured and unstructured—to build real-time risk scores. By recognizing subtle links between counterparties across systems, it reduces false negatives and highlights nuanced or complex risks.

2. Evolving Sanctions and Regulatory Frameworks

The problem: Sanctions lists change frequently. Keeping up with updates from OFAC, EU authorities, and regional regulators creates ongoing friction—especially when processing payments from or through high-risk markets like Mexico or Kenya.

Recent example: In March 2025, OFAC added over 200 new entities across crypto and banking sectors to its Specially Designated Nationals (SDN) list—a major spike compared to previous years.

Where Cognitive AI helps: Cognitive AI platforms integrate real-time sanctions screening with transaction monitoring, flagging not just direct matches but also fuzzy linkages (e.g., intermediary banks with indirect ties to sanctioned entities). More importantly, it adapts faster than static list-matching systems, helping compliance teams stay in line with regulatory updates.

3. Customer Friction vs. Compliance Trade-Off

The problem: Customers expect frictionless payments, whether domestic or cross-border. These expectations can’t always be met with compliance requirements, and legacy rule-based systems have often caused issues with their thresholds being met, requiring organizations to conduct manual due diligence on the customer and/or transaction before allowing the payment to proceed. This can be especially the case when payments are going to high-risk jurisdictions, causing significant delays. 

Why it matters: Causing unnecessary delays in payments creates friction and frustration for customers. This can lead to customer attrition where they seek out alternative payment processors where they can be provided with an easier system to make their transactions. 

Where AI helps: AI enables smarter, adaptive transaction screening that supports faster domestic and cross-border processing. By assesing risk in context, not just against fixed rules, it reduces the chance of transactions being unnecessarily delayed or flagged. The result: stronger ocmpliance without compromising speed or customer experience.   

4. Hidden Network Risks

The problem: Criminal networks increasingly exploit gaps in correspondent banking chains, layering transactions across multiple jurisdictions to evade detection.

Case in point: FATF’s 2021 report on Trade-Based Money Laundering Risk Indicators highlights the growing role of opaque payment corridors. Key risk indicators include inconsistent documentation, usually complex payment structures, and transactions lacking clear economic rationale. 

Where Cognitive AI helps: By continuously learning from transaction patterns, Cognitive AI identifies suspicious activities across entire payment networks, not just isolated transactions. This “whole-network” visibility enables institutions to detect red flags such as value discrepancies, unusual shipping routes, and frequent round-dollar transfers. Such indicators align with FAFT’s published risk typologies, helping compliance teams intervene earlier and more effectively. 

5. High False Positive Rates

The problem: In cross-border payment monitoring, it’s not unusual to see 95–99% of alerts classified as false positives. That’s operationally unsustainable—especially for fintechs scaling lean teams.

Industry perspective: According to Datos Insights 2024 AML professionals survey, most AML models that financial institutions use routinely generate 90% to 95% false positive rates, which most compliance teams cite as their top pain point in cross-border payments.

Where Cognitive AI helps: By using adaptive learning models rather than static rule engines, Cognitive AI can dynamically recalibrate what constitutes normal versus suspicious behavior. ThetaRay’s platform, for instance, applies Cognitive AI logic that can understand the context and nuances behind customer behavior and transaction – clearly identifying what is high risk and what is normal behavior–minimizing noise while maximizing detection effectiveness.

Aligning technology with compliance expectations

It’s worth emphasizing: regulators aren’t just looking for technology adoption. They expect financial institutions to choose the right technologies.

• • FATF’s Suggested Actions to Support the Use of New Technologies for AML/CFT highlight the importance of responsible innovation in AML/CFT. This includes ensuring transparency, explainability, privacy safeguards, strong cybersecurity, human oversight, and alignment with international standards. Cognitive AI solutions that integrate these principles support risk-based frameworks while enhancing financial inclusion and effectiveness.

• • The UK’s FCA and U.S’s FinCEN both published updated advisories encouraging financial institutions to leverage machine learning models, while ensuring proper governance structures.

Choosing solutions that offer explainability, audit trails, and compliance-aligned controls isn’t just good practice. It’s becoming a regulatory expectation.

How to shift compliance from cost center to strategic advantage

Historically, AML and compliance functions in cross-border payments have been seen as cost centers—necessary but burdensome.

That mindset is shifting. Financial institutions that invest in adaptive technologies like Cognitive AI aren’t just protecting themselves from fines. They’re building faster, safer, and more trusted cross-border payment services that customers value.

In a space where speed, trust, and compliance intersect, that’s not just a defensive play. It’s a smarter growth strategy.