IF AI AND MACHINE-LEARNING-EQUIPPED PLATFORMS ARE THE ULTIMATE TBML SOLUTION FOR COMPLIANCE, WHERE ARE THE REPORTS OF SUCCESSFUL IDENTIFICATION AND INTERDICTION OUTCOMES?

Lately, there have been a number of articles, some of them little more than thinly-disguised infomercials, touting the virtues of Artificial Intelligence and machine learning-enabled programs for compliance officers targeting Trade-Based Money Laundering in their transaction monitoring assignments. These offerings, which claim that they effectively combat TBML, reduce false positives, and employ holistic analysis to correctly ferret out actual risks, have not resulted in a sea change, by my reckoning.

I am all for any innovative advancement in TBML identification, especially those that have moved past a rules-based approach, to effectively approach human analysis of data, but if such a platform exists, and is in general use, why have we not seen or heard the results? Certainly, if such systems are in the hands of compliance officers at the world's larger financial institutions that have a major foreign trade clientele, such good news would have already permeated the TBML compliance world; there would have been media coverage of law enforcement arrests of laundrymen, and the dismantling of their networks.

Instead, we have heard none about such vaunted AML successes; compliance officers have not widely shared with their colleagues the names of such programs, and they have not surfaced at conferences, seminars and events. This lead me to believe that what I am reading is little more than the employment of marketing skills, either to sell a specific product, or to jump wholeheartedly on the AI bandwagon, to appear up-to-date on a potential solution to the compliance profession's TBML failures.

Perhaps such a program will one day come along, but in 2024, the only effective TBML solution, is in my humble opinion:
(1) Having comprehensive knowledge of Trade-Based Money Laundering tradecraft, how its laundrymen create new typologies, and an understanding of their internal culture.
(2) A dataset of records having a global scope, on international trade transactions, so that individuals conducting transaction monitoring have all possible reference points, from prior payment history.
(3) An analytical approach, performed by compliance officers with both cultural literacy and broad-based knowledge of the world of legitimate business.

Until that perfect AI-powered platform is built, tested and passes scrutiny, catching TBML in real-time happens only when a compliance officer who has the right stuff, and with the right data, analyzes a transaction, and correctly deduces what he is seeing on his screen is money laundering.