In a bright publication from University of Chicago, Mirko Nazzari and Peter Reuter have recently provided a critical evaluation of the global AML regime.
Their main conclusions highlight that, despite decades of regulation, vast sums spent, increasing sophistication in rules, and compliance burdens, there is no strong evidence that money laundering is declining, or that the laundering of criminal proceeds has become meaningfully more difficult or expensive.
Global weaknesses of AML/CFT systems
Some of the key identified weaknesses of the AML/CFT systems include:
Over‐emphasis on compliance form over substance
Many institutions do what is needed to satisfy regulators—filing Suspicious Activity Reports (SARs), implementing thresholds, etc.—but these actions often don’t translate into actual disruption of laundering activity.
The system tends toward behaviors like “defensive filing” (filing a lot of reports just to avoid liability), rather than targeting those likely to be meaningful.
Regulations growing complex but effectiveness not keeping pace
Rules, standards, supervisory expectations, and regulatory frameworks have become more elaborate. But enforcement, oversight, and the capacity to respond often lag behind. Resources, institutional capacity, data quality, operational skills, and technological tools are often insufficient.
High false positives, inefficient alert systems, and cost inflation
Many AML tools produce large numbers of alerts, most of which turn out to be false positives. This overloads compliance teams, creates high operational cost, and can dilute attention on genuinely risky cases.
Lack of adequate metrics for effectiveness
There is too much focus on process metrics (number of reports filed, structure in place, etc.) and too little on outcome metrics (how much money was actually prevented from being laundered; how many criminal networks disrupted). It’s hard to assess whether AML/CFT systems are working in their substantive mission.
Regulatory fragmentation and variation in capacity across jurisdictions
Differences in laws, supervision, enforcement, data availability (e.g., beneficial ownership), institutional infrastructure, and resources mean effectiveness varies dramatically across countries.
Some jurisdictions struggle to even implement basic AML/CFT measures or put in place regulations without the capacity for appropriate control or sanctions.
Evolving criminal/technical sophistication
Launderers adapt. They exploit gaps: cross-border flows, informal financial systems, new technologies (crypto, virtual assets), money mules, shell companies, etc.
Legacy systems are often not agile enough to detect new typologies.
Balance between AML compliance & other goals
Business development, financial inclusion, privacy, customer experience, cost efficiency—all impose friction or trade‐offs. Meeting regulatory demands often means increasing friction for legitimate customers or imposing high costs on regulated entities—sometimes beyond what they can sustain.
Why these challenges persist: structural & operational roots
These problems are not just “bugs,” they arise from deeper structural issues.
Resistance or inertia
Regulatory bodies, financial institutions often prefer incremental improvements rather than radical reforms, partly for fear of exposing themselves to risks (regulatory, reputational).
Lack of feedback loops
Many systems do not measure or learn from enforcement outcomes. Which alerts led to arrests / convictions? Which didn’t? There’s little systematic learning.
Data silos & poor interoperability
Institutions may not share data; cross‐institution, cross‐jurisdiction intelligence is largely underutilized.
Legacy technology & rule‐based systems
Many systems still rely heavily on fixed thresholds, static rules, or manual reviews, which don’t scale or adapt well.
Resource constraints
Staffing, skills, budgets, training, technological investments are uneven, especially in smaller institutions or in less wealthy jurisdictions.
Heptalytics addresses these challenges and give an innovative boost to compliance
Heptalytics offer many of these features to fill the gaps identified by Nazzari & Reuter, and help to move the AML/CFT system toward greater substantive effectiveness
Advanced AI & graph/neural network analytics
Graph/network‐based models can identify relationships that transaction‐threshold rules miss: money mules, layering, linkages among accounts. This helps detect more complex criminal patterns.
With typical rule systems, many connections are invisible; graphs & embeddings help bring those to light. Heptalytics shifts the paradigm from rule‐based threshold triggers to contextual, relational, behavioral detection. This is more in line with emerging “next generation” AML thinking — while respecting regulatory constraints, focus is given to substance and outcomes.
Business Pattern Library & detection of new/novel patterns
Patterns avoid being stuck with customized thresholds-rules. Criminals often evolve typologies; having a library of patterns, being able to test and back‐test novel ones, can help stay ahead of the curve.
Heptalytics natively uses feedback loops: back‐testing, management dashboards, alerts prioritization—all help institutions observe what is working, what isn’t, then refine their approaches.
Reducing false positives & alert overload
Since false positives are a key inefficiency (wasting resources, creating alert fatigue), reducing false positives means compliance staff can focus on high‐risk alerts, so much better outcomes and lower cost.
By reducing false positives dramatically, Heptalytics frees up resources for deeper investigations. That helps with the “alert fatigue” issue, and means better signals can get attention, rather than being swamped by noise.
Prioritisation & decision support
Not all alerts are equally risky. Prioritizing based on risk, connecting cases to account level, enabling operations to know where to allocate attention can improve efficiency and effectiveness.
Data science as a service / context‐adapted modelling
Instead of generic tools, being able to adapt the model to each institution’s context—its customers, geography, business model, risk profile—improves relevance. Also, deploying data science experts reduces operational burden.
Monitoring & reporting tools for oversight & management
For effectiveness, management needs metrics and dashboards to track performance. They also help with auditability and regulatory reporting.
Complementarity with other systems
Legacy systems (authorization engines, rule‐based modules, etc.) are not replaced but augmented. That allows smoother integration, less disruption and better overall performance.
Conclusion
The core message from Nazzari & Reuter is that the global AML/CFT system has matured in terms of regulation, process, cost—but not necessarily in terms of effectiveness. To move to the next level requires better tools, better metrics, better data, and systems that can adapt, focus, and scale.
Heptalytics appears to align with many of the directions that theory, academic critique, and regulatory suggestions are pointing toward: using advanced analytics, reducing false positives, enabling better prioritization, giving compliance teams more intelligent decision support, tailored modeling, etc.
