Artificial Intelligence in Money Laundering Prevention

Can artificial intelligence help prevent money laundering? In this webinar, Víctor Salinas — an ACAMS-certified anti-money-laundering specialist with experience in Ecuador, Colombia and Peru — uses real examples to show how a compliance officer can rely on AI to detect red flags, and why human analysis remains indispensable: AI generates alerts, but an alert alone is not enough to deny a service or file a report; someone has to analyze it.

Four uses of AI in AML/CTF prevention

• Detecting suspicious patterns. Machine-learning algorithms find what goes unnoticed by hand, such as smurfing or structuring: many small transactions that, taken together, add up to significant amounts in order to evade reporting thresholds. Well configured, they also reduce false positives. This use demands data: without clean databases of customers, suppliers and employees, the tool ends up overrated.
• Knowing your counterparties. AI keeps profiles up to date and segments customers by homogeneous behavior, raising a flag over whoever departs from their usual pattern.
• Analyzing unstructured data. News, press releases and social media: AI summarizes what the web holds about an individual or a company — as long as the information is public — and cites its sources.
• Working with the specialists. AI brings efficiency; the compliance officer brings the judgment to decide whether to start or continue a business relationship, or to file a suspicious transaction report (STR).

How to write a good prompt

The tool — Copilot, Gemini or ChatGPT — matters less than the instruction. A good prompt is relevant, adaptable, original and clear: include keywords such as “money laundering”, “acts of corruption” or “criminal offenses”; state the sources to analyze (news, press releases, social media); use full names, plus the nationality for foreign nationals, and narrow the search to periods of about ten years. Always ask for the sources behind each finding so you can document your report: a bulletin from the Attorney General’s Office carries more weight than a single media outlet. The video walks through three ready-to-copy prompts with real results: a company with confirmed links, its transactional analysis, and a case with no findings.

Bring this efficiency to your due diligence

AI helps you search; the decision is still yours. To automate the screening of national and international watchlists with evidence taken straight from the sources, meet AplaSoft, our due diligence software, or request a free demo with our specialists.