Trade-based financial crime (TBFC) exploits the complexity of global trade, slipping past outdated compliance systems. In this Q&A, Eastnets’ Hassan Zebdeh discusses why banks struggle to detect TBFC, how AI can help, and what’s needed to close the gap.
Trade-based financial crime (TBFC) is one of the biggest blind spots in global finance. Unlike fraud or money laundering, which have well-established detection frameworks, TBFC thrives in the complexity of cross-border trade, slipping through the cracks of traditional compliance systems. The numbers are staggering—$1.6 trillion in illicit transactions annually, yet just 0.2% of Suspicious Activity Reports (SARs) explicitly cite TBFC.
Despite its scale, many financial institutions still rely on outdated methods to detect and prevent these crimes. The complexity of global trade—multiple parties, jurisdictions, and fragmented documentation—makes it easier for bad actors to manipulate transactions. And while AI is transforming fraud detection and anti-money laundering (AML) efforts, its adoption in TBFC remains slow. Experts warn that financial institutions must act decisively before criminals widen their technological advantage even further.
To explore why banks are facing these challenges and how technology can play a role in combating TBFC, we spoke with Hassan Zebdeh, Financial Crime Advisor at Eastnets, who shared insights into the risks, detection gaps, and the role of AI in modernizing financial crime prevention.
The Scale and Complexity of TBFC
Trade-based financial crime takes advantage of the sheer volume of global trade to obscure illicit transactions. According to Zebdeh, 30% of all global money laundering cases are linked to trade finance, yet the industry lacks a unified system to track and report these crimes effectively.
“Trade-based financial crime is inherently complex,” he explains. “It involves multiple parties, layers of documentation, and jurisdictions.”
Criminals exploit mis-invoicing, ghost shipping, and price manipulation to disguise illicit financial flows. Ordinary pens valued at hundreds of dollars or bulldozers drastically undervalued by tens of thousands are just some of the ways bad actors distort financial records to launder money.
The challenge lies in visibility. With $32 trillion in trade transactions annually, banks only see a fragmented portion of the deal, making it difficult to assess risk holistically. “Each bank only sees part of the transaction, but there is no industry-wide system that allows them to evaluate the full risk,” Zebdeh notes.
How AI Is Changing the Landscape
Traditional methods for detecting TBFC have struggled to keep up with the scale and complexity of modern global trade. AI is now offering a way forward, helping financial institutions analyze vast amounts of trade data, flag anomalies, and automate previously manual processes.
One major breakthrough is in document digitization. Trade-based transactions generate huge volumes of unstructured data—bills of lading, invoices, contracts—many of which are still paper-based. AI-driven Optical Character Recognition (OCR) tools have dramatically improved accuracy rates, allowing banks to digitize and analyze trade documents with 95%+ accuracy.
AI is also helping flag discrepancies in trade values, identifying pricing anomalies that may indicate fraudulent activity. “For example, if a shipment invoice lists a standard office chair as costing $5,000 per unit—far beyond its market value—the system can flag it for review,” explains Zebdeh. By cross-referencing historical pricing data, trade classifications, and transaction records, AI can pinpoint irregularities that warrant deeper investigation.
AI-powered vessel and route monitoring is another emerging solution. Most financial institutions lack real-time visibility into how goods move across borders. “If a shipment’s route unexpectedly changes or passes through a sanctioned area, AI can flag this as a potential risk,” Zebdeh says.
The Roadblocks to AI Adoption
Despite AI’s potential, banks have struggled to implement these technologies at scale. Adoption remains inconsistent, with many institutions relying on outdated IT systems, siloed data, and legacy compliance frameworks.
“Trade documents are rarely standardized,” Zebdeh explains. “Banks often use inconsistent formats, making it difficult to consolidate and analyze information effectively.” This fragmentation limits AI’s ability to provide a comprehensive risk assessment, particularly when transactions span multiple jurisdictions and involve numerous counterparties.
Another issue is the high volume of false positives in existing transaction monitoring systems. Many compliance teams are overwhelmed by alerts that lack contextual analysis. “Most systems still flag transactions based on predefined thresholds rather than broader, AI-driven pattern recognition,” says Zebdeh. This creates a huge operational burden for financial crime teams, making it harder to identify genuine risks.
Regulatory uncertainty is also a key barrier. While regulators like the Financial Conduct Authority (FCA) have urged financial institutions to adopt continuous monitoring for trade transactions, there is no universal mandate for AI-based detection. Without clear regulatory incentives, many banks remain hesitant to invest in AI solutions.
A Path Forward
The coming years will likely see greater AI adoption in TBFC detection, but the industry must move faster. Emerging solutions, such as electronic bills of lading and standardized e-invoices, could prevent duplicate or fraudulent claims and enhance transparency in trade finance. However, adoption remains inconsistent, and there is still no shared, industry-wide screening system.
Zebdeh highlights the risks of inaction, pointing to recent cases where luxury goods were rerouted through countries bordering Russia to evade sanctions. “This highlights how trade flows can mirror broader geopolitical and economic shifts,” he explains. “AI systems with real-time monitoring capabilities are essential for detecting these patterns.”
Collaboration between regulators, financial institutions, and technology providers will be critical in strengthening TBFC detection. While bodies like the Financial Action Task Force (FATF) and Wolfsberg Group have issued global recommendations, practical implementation remains a challenge.
Ultimately, AI alone won’t solve TBFC. Banks need to modernize their compliance infrastructure, break down data silos, and improve intelligence-sharing across jurisdictions. Without a more coordinated approach, financial institutions risk falling further behind criminals who are already leveraging advanced technology to exploit gaps in trade finance.
Leave a Reply