By: Tim Staheli

Global trade in goods and commercial services reached $34.65 trillion in 2025. Every one of those transactions generates paperwork. Banks process those documents, check them against international rules, and release funds or guarantees accordingly. The people responsible for that checking are skilled, trained, and (relative to the task) comprehensively outnumbered. Data is scattered across banks, shippers, insurers, and customs authorities, and the detailed documentation that accompanies every transaction creates a presumption of legitimacy that is, in practice, difficult to refute.

Buried in that stack of documents, however – a bill of lading, a commercial invoice, a certificate of origin – there is a lie. It might be a price that bears no relation to market rates. It might be a shipment of consumer electronics that, on closer inspection, was never shipped. It might be a company with a registered address and a bank account, but no discernible logistics footprint. The deceit is not sophisticated. There is just too much to check.

In this fog, trade-based money laundering (TBML) can run riot. And the tricks are disarmingly simple. Over-invoice a shipment, and the buyer transfers more money than the goods are worth, moving value across borders under commercial cover. Under-invoice it, and the seller repatriates funds in plain sight, below the threshold of suspicion. Or invent the shipment entirely.  Conjure the paperwork, forge the bills of lading, and pocket the difference on goods that never existed. Cash-based money laundering, by contrast, is noisy. It trips alarms. TBML hides in the paperwork. To a tired pair of eyes working through a hundred transactions on a Thursday afternoon, all documents begin to look normal.

One case, documented in the Monetary Authority of Singapore’s 2024 Money Laundering Risk Assessment Report, tells the story well. Ng Kheng Wah, director of T-Specialist International, had been shipping luxury goods – wines, spirits, jewellery, watches – to a North Korean department store for the better part of a decade. When the North Korean buyer fell behind on payments, Ng came up with a solution: he produced 81 fictitious invoices from a shell company, purporting to document the sale of Watari-branded instant noodles, submitted to five Singaporean banks to unlock over $95 million in trade financing loans. The noodles didn’t exist, and neither did the transactions. Five banks, each processing what looked like routine trade paperwork, handed over the money.

The fentanyl trade provides an even harsher illustration of what is at stake. A FinCEN report published in April 2025 detailed complex Chinese money-laundering schemes that routinely involved electronics – mobile phones, vaping devices – used to layer and move cartel proceeds. The goods were real enough, shipped and invoiced like any legitimate consignment. But the invoices were manipulated, the prices fictitious, and the money moving through them was fentanyl proceeds. Between 2020 and 2024, Chinese networks and Mexican cartels laundered more than $312 billion through TBML schemes, according to a FinCEN advisory. The trade finance system did not only witness that figure. It enabled it.

FATF’s most-cited estimate puts TBML at roughly $1.6 trillion annually, a number some analysts consider an undercount, with Financier Worldwide suggesting the true figure may be closer to $2 trillion. Whatever the truth, the trajectory is hardly improving. Rising tariffs have increased the incentive to manipulate invoice values. Sanctions regimes have pushed sanctioned entities to exploit trade channels as an alternative to the now-scrutinised correspondent banking network. The shadow fleet servicing Russian oil exports is, among other things, a TBML story.

Against this backdrop, the response from major banks has for years been essentially the same: more staff, more training, more suspicious transaction reports. No document checker is fast enough. For years, banks acknowledged as much and carried on regardless.

Fortunately, that is beginning to change. NatWest’s recently announced partnership with Cleareye.ai, the US-based AI platform built for trade finance, is one of the cleaner illustrations of where the industry is heading. Under the agreement, NatWest will deploy ClearTrade® across its commercial and institutional trade operations, automating document extraction and classification, running ICC rule-based examinations, and performing TBML compliance checks: everything a trade finance compliance team does, rendered in software.

Michael Gilham, Trade Product Lead, Commercial and Institutional at NatWest, said: “As Britain’s biggest business bank, we’re focused on giving customers the confidence to seize growth opportunities. This partnership will help them trade in foreign markets with greater speed and certainty, while enabling our colleagues to deliver a more personalised and productive service.

“It will also strengthen protection against fraud and financial crime, by using innovative technology to enable us to provide better service to our customers.

“As we embrace leading, innovative technologies, we remain committed to harnessing them responsibly enhancing customer experiences while strengthening our risk and control environment. At every stage, we will prioritise the safety, trust, and wellbeing of our customers and colleagues.”

Cleareye.ai is not a new name in this space. JPMorgan, Bank of America, Mizuho, Riyad Bank, Gulf International Bank and Lloyds, which signed a comparable deal in 2024, are all ClearTrade® clients. In an industry not known for rushing towards new technology, that kind of roster tends to make the due diligence argument for itself.

Mariya George, Chief Executive Officer and Cofounder of Cleareye.ai, added, “NatWest is leading the way in reimagining trade finance, and we are proud to support their innovation journey. ClearTrade® is purpose built to deliver both automation and compliance outcomes, and we are excited to help NatWest unlock scalable impact for their teams and their clients.

“This partnership highlights NatWest’s ongoing focus on building operational resilience, embracing advanced technologies, and staying ahead of regulatory demands in the complex global trade environment. It reflects the bank’s continued commitment to innovation that delivers meaningful outcomes for clients and the broader financial ecosystem.AI in compliance settings carries a particular anxiety: the black-box problem. A system that cannot show its reasoning has no business making compliance decisions. Here, transparency is not optional.”

Ultimately, the trade finance industry has a volume problem that no hiring plan could ever solve. The volume of trade, and of the documents that underpin it, will not decrease. The obligation to look more closely at those documents, as TBML figures grow and regulatory patience narrows, will not be relaxed. The question was not whether the manual model was adequate. It was always how long it would take for enough people to admit that it wasn’t. For NatWest, that admission has arrived – and it comes with a solution.

The fog, at last, is beginning to lift.