Standard Chartered prevented ~$90M from being laundered by helping investigators surface higher-risk accounts faster

As a leading international bank connecting clients across some of the world’s most dynamic markets, Standard Chartered sees money move at extraordinary scale, and takes seriously its role in protecting those flows. Across a billion transactions, criminal activity is designed to hide in plain sight. The bank built an AI-powered system that helps investigators identify the signals that matter before the money disappears.

A billion transactions. And criminals trying not to be noticed.

Money mule networks can move illicit funds through personal and business accounts in ways designed to evade traditional transaction-monitoring thresholds.

For Standard Chartered's Financial Crime Compliance teams, the signals were there, transaction patterns, cross-border activity, customer history, digital behavior, and device intelligence. But they lived across different systems, leaving investigators to piece together the picture before they could determine which accounts deserved attention.

Standard Chartered didn't need more alerts. It needed to know which ones mattered.

Changing the starting point

Standard Chartered brought financial crime specialists, data scientists, and engineers together to build STABLES, an AI-powered scoring engine that prioritizes money mule risk. Every day, STABLES combines behavioral signals across the bank and ranks the accounts most likely to require action, showing investigators the behaviors behind each score. Instead of starting with an undifferentiated queue, investigators start with the accounts most likely to require action.

The difference isn't more alerts. It's knowing where to look first.

40,000 accounts surfaced. ~$90M stopped.

In the first 12 months after deployment, STABLES surfaced approximately 40,000 accounts for investigation. Mean investigation time fell approximately 75%, turning work that could take months into a matter of days. Investigators can review roughly three times as many alerts.

Most importantly, Standard Chartered froze or prevented approximately $90 million from being laundered before criminal networks could cash out. And another important signal came from outside the bank: external police alerts related to money mule activity declined approximately 50%, indicating that Standard Chartered was becoming a less attractive target for money mule networks.

AI built on human expertise

STABLES doesn't replace the people who understand financial crime. It scales what they know. AML and Financial Crime Compliance specialists helped translate years of expertise about money mule behavior into the way risk is assessed. 

Dataiku brings together data from banking, customer, digital, and device systems; coordinates feature engineering and daily model scoring; and delivers prioritized results into investigators' existing environment. Crucially, investigators can see the behaviors behind each score. AI identifies and prioritizes the risk. Investigators decide what to do about it. The same foundation also allows Standard Chartered to adapt STABLES across markets and both retail and corporate banking without rebuilding the system each time.

As financial crime evolves, specialists can refine the risk logic, model teams can test and govern changes, and investigators remain responsible for the final decision. What began as a better way to find higher-risk accounts has become a repeatable way for Standard Chartered to turn financial crime expertise into action across markets, customer segments, and billions of transactions.