From Detection to Defensibility: Malaysia, Catch Me If You Can: Teaching AI to Think Like a Criminal
Following the success of Flagright’s first financial crime handbook, the new Global Handbook for AI in Financial Crime Compliance asks how far AI can go in fighting financial crime, and how you defend every decision it makes.
Malaysia has never needed the answer more. Online commercial crime nearly doubled in 2025, with almost RM3 billion lost. Police flagged more than 50,000 mule accounts in 2024. RMiT already requires behavioural risk profiles for every customer and checks on transfers to suspected mules.
But by the time money reaches a mule, the scam has already worked. Every mule was a normal account once, and criminals plan the whole chain, from the scam that takes the money to the accounts that move it. What could AI catch if it thought like a criminal?
To help answer that, we are joined by Malaysia’s leading crime prevention expert, Dr. Shamir Rajadurai, founder of Prevent Crime Now, who will open the morning with a talk on how criminals actually plan and run these operations, followed by a fireside chat on what that means for catching mules early.
Join fraud, AML, and compliance leaders from Malaysian financial institutions for breakfast, a panel, and a live demo of AI catching a mule.
Agenda
7:30 AM: Registration and coffee refreshments
8:30 AM: Panel discussion
9:15 AM: Live demo: catching a mule before it moves money
9:30 AM: Brunch and networking
11:00 AM: Close
Panel Themes
Spotting a mule before the first dirty ringgit lands, even when the account holder doesn’t know they’re one
Scam and laundering are one operation. Why do banks still split them across fraud and AML?
Turning what criminals reveal in prison interviews and investigations into signals a model can watch
Acting on an early flag without punishing innocent customers
Only a quarter of practitioners trust AI enough to act on it. What would change that?