Anti-money laundering in transition: what's coming with AMLA – and how AI can help

Every year, an estimated two to five percent of global gross domestic product is laundered through money laundering, amounting up to 2 trillion US dollars. In Germany alone, the figure is believed to be around 100 billion euros annually. The consequences are devastating: money laundering threatens the integrity of the financial system, distorts competition, and strengthens criminal networks.

How will the new EU Anti-Money Laundering Authority (AMLA) intensify the fight against money laundering – and what part will AI play? This article shows how technology and regulation will work together in the future, and what financial institutions should be ready for.

How the EU is reshaping its legal framework

The European Union has begun a fundamental overhaul of its legal framework to combat money laundering and terrorist financing. From 2027, a cash limit of 10,000 euros is to apply across the EU in order to curb anonymous transactions. In addition, the Anti-Money Laundering Authority (AMLA) has been created as a central EU anti-money laundering authority.

Starting from summer 2025, it will be responsible for the strategic management and coordination of national supervisory authorities, particularly in cross-border cases. As part of its duties, it will develop technical standards and procedures to ensure uniform Europe-wide requirements for combating money laundering, terrorist financing, and the application of financial sanctions.

Increasing pressure on financial institutions

These new guidelines and requirements will significantly heighten the pressure on financial institutions to act. With each new rule, it becomes clearer to financial institutions that the requirements for processes, documentation, and risk assessment in the prevention of money laundering are constantly increasing.

Banks and payment service providers are already taking on key tasks in the fight against financial crime: The German Financial Supervisory Authority BaFin reports that over 90% of the 320,000 suspicious activity reports in 2023 came from the financial sector.

At the same time, many institutions work with historically evolved IT landscapes in which KYC checks, transaction monitoring, and sanctions list comparisons run separately. This fragmentation makes it challenging to obtain a complete risk picture, especially for complex, cross-border cash flows.

Added to this the sheer volume of data: Millions of transactions pass through systems every day, making it enormously time-consuming to correctly assess anomalies. Many compliance teams struggle with a flood of cases and a high rate of false-positive alerts that have to be checked manually.

So how can these challenges be overcome? The financial sector needs new approaches to master the balancing act between stricter supervision and limited resources.

AI in the prevention of money laundering

The potential:

This is where modern technology, such as Artificial intelligence (AI) comes into play as a key technology because conventional, purely rule-based systems are reaching their limits. AI-supported processes can recognize patterns faster and more precisely within huge amounts of data. Such intelligent systems learn from past cases, adapt to new money laundering methods, and significantly reduce false alarms.

According to a McKinsey study, AI can increase the detection rate of suspicious transactions by up to 30% – a major advance in the face of limited compliance teams.

In addition, modern AI platforms offer structural advantages: they bring together data from various compliance areas on one common platform (from customer profiles and transaction histories to screening results). This creates a consistent, dynamically updated risk profile for each customer.

Hybrid approaches consisting of data-driven models and knowledge-based checks have proven particularly successful. In addition to providing the necessary traceability and explainability to ensure that AI decisions are accepted both internally and externally, they deliver better results. In this way, AI can be seamlessly integrated into existing control and governance structures, an important prerequisite for use in regulated institutions.

Hurdles:

Despite this potential, many institutions are still hesitant. Regulatory uncertainty is a key obstacle. Supervisory authorities have so far been reluctant to accept black box AI models, especially if their decision-making processes are difficult to understand. Many banks therefore lack guidance on how AI can be reconciled with applicable compliance requirements.

Clear standards on the use of AI in AML processes are largely lacking: questions regarding explainability, documentation requirements, or the permissibility of certain model types are often unanswered. Out of caution, many financial institutions therefore prefer to stick with established but less powerful systems instead of implementing innovative AI solutions.

AMLA as a trailblazer for innovation

The new EU authority AMLA could resolve these blockades. As a central authority, it should not only directly supervise selected high-risk institutions. Above all, it should develop and enforce uniform technical standards. A harmonized, data-driven regulatory approach would reduce legal uncertainties in the use of AI and provide institutions with clear guidelines for the use of new technologies.

AMLA has the potential to significantly accelerate the introduction of AI in money laundering prevention: Consistent rules and the active promotion of technological innovation can reduce existing hurdles.

Institutions that are early adopters of AI could thus further extend their lead if AMLA creates a binding framework for everyone. Whether the AMLA will actually become the hoped-for catalyst for technological progress in the fight against money laundering remains to be seen. But the potential is recognizable.

Conclusion and outlook

The introduction of the AMLA does not mark a completely new start, but it does underline the existing trend: regulatory pressure continues to increase. For banks and payment service providers, it is no longer a question of whether they need to adapt their processes, but how quickly and specifically this can be done.

Artificial intelligence is no longer a topic for the future. It is a concrete lever for increasing efficiency and quality in money laundering prevention. AI helps to identify risks earlier, streamline processes, and meet the increasing supervisory requirements.