AML is Fighting Yesterday's War

Every major anti-money laundering (AML) framework is built on yesterday's reports. Often, by the time typologies become guidance, criminals have already adapted and moved on to newer and more sophisticated ways of perpetrating fraud.

For a very long time, regulators have been telling financial institutions that the separation of anti-money laundering, fraud, cybersecurity and compliance teams is no longer optimal, as most financial crime starts as fraud and is interconnected with cyber breaches. In 2016 and 2019, the Financial Crimes Enforcement Network (FINCEN) issued advisories urging financial institutions to collaborate across teams, as siloed work creates vulnerabilities.

Almost a decade later, this advice still sounds like a dream that is yet to be actualized. Most money laundering and fraud are now AI-enabled and faster than a monthly, quarterly or even yearly transaction monitoring review can register. The problem that most anti-fraud or financial crime specialists face is not a matter of program efficiency but instead that their capabilities were designed for a threat landscape that has fundamentally evolved. 

Moving from Detection to Anticipation 

The traditional AML framework is fundamentally retrospective. Financial institutions' transaction monitoring systems are configured to flag transactions based on known typologies or previous events, after which investigators review the alerts to determine if they are suspicious transactions that require a Suspicious Activity Report (SAR) filing. The time lapsed between these events may often be plenty for the money to be moved by the fraudster.

In a case where the activity is not a known typology, the transaction monitoring system is unable to detect it, limiting its effectiveness and the investigators’ ability to intervene.

Feedzai's 2026 compliance predictions describe a shift from pattern recognition to anticipatory systems. This means that financial programs will not be built to just catch what has already happened but flag what is likely to happen next. Most financial institutions are lagging here: The NICE Actimize “AML Tech Barometer 2026” shows that only 30% of institutions currently treat AML-fraud integration as a business priority.

The challenge facing anti-fraud professionals today is not just that criminals have become more sophisticated, but that these perpetrators are operating on a different timeline.

Currently, it takes criminals several days to test new schemes and perfect them, while financial institutions spend months updating their programs or fraud typologies. This creates a gap. As one side continuously adapts and develops new trends, the other responds to these trends only periodically. No matter how effective the traditional program is, it will struggle to keep up with the changing activities of fraudsters. 

Where the Traditional Framework Fails 

Three trends illustrate how quickly things have changed. 

1. The increase in synthetic identity fraud: What previously required real, stolen documents can be manufactured quickly, convincingly and remotely with the use of generative tools and onboarding fraud built using synthetic or AI-fabricated identities. Deepfake-enabled attacks have now become a regular occurrence. Entrust has estimated that a deepfake-driven attack occurs somewhere in the world every five minutes. 
2. Real-time payments: The European Banking Authority identified that fraud risk on instant payments is roughly 10 times higher than fraud on traditional transfers, primarily because there is no longer a window for settlement of transactions in which suspicious transactions can be caught, reviewed and reversed before the transfer is completed. 
3. Converging threats: Cybercrime and money laundering have become different manifestations of the same problem. Business email compromise, account take-over and money laundering operations are mainly perpetrated by the same actors through the same structure. Monitoring them individually means that each team only gets to see one part of the much larger picture. 

These trends point to a broader reality: Incremental improvements to traditional financial programs will not be enough.  

A Framework for Adaptive Financial Crime Programs 

Since financial crime has become adaptive, it is paramount for the financial crime program to become adaptive, too. This requires new technologies and rethinking how institutions should identify, prioritize and respond to emerging fraud threats.

Below are four principles that can guide institutions on how to adapt to this change: 

• Anticipate: Move from relying solely on historical typologies by continuously monitoring emerging fraud patterns, payment innovations and criminal methods before they become a common occurrence. 
• Integrate: Treat fraud, money laundering and cybersecurity as interconnected risks, not separate functions. Criminals operate across all three, and a unified approach improves early detection. 
• Adapt: Change static control environments to models that allow rules, risk indicators and investigative priorities to evolve as quickly as possible. 
• Learn: Develop feedback loops that identify lessons learned from fraud attempts, investigations and regulatory development to improve continuously rather than during scheduled reviews.

None of this means abandoning existing financial crime programs. It means questioning long-held assumptions about how financial crime is organized, detected and investigated, and building programs that reflect how criminals actually operate today. That could include shortening review cycles, creating an integrated financial crime governance, testing for future scenarios and measuring effectiveness in conjunction with activity.

These changes should work to ensure compliance remains effective in an environment where financial crime is constantly evolving. The question for today’s fraud examiners is no longer: “Are we compliant?” It is: “Are we prepared for tomorrow’s threats?” The future of financial crime will belong to institutions that learn and adapt as quickly as the criminals they seek to stop.