How can AI support with anti-money laundering?

The formation of the Anti-Money Laundering and Countering the Financing of Terrorism Authority (AMLA), along with legislative changes designed to create a unified EU anti-money laundering (AML) and counter-terrorism financing (CFT) rulebook, presents a unique opportunity to strengthen the AML and CTF framework.

AI-enabled solutions can significantly enhance AMLA’s efficiency, allowing regulated firms and supervisors to address high-risk issues more quickly, automate time-intensive tasks, and maximise the productivity of their teams. This will enable AMLA to perform its supervisory and enforcement role more effectively, in a cost-efficient manner.

However, safeguards need to be in place when incorporating AI, especially in light of the EU’s Artificial Intelligence Act (AI Act), which governs the deployment of AI systems across sectors. Factoring in new and advanced technologies early in AMLA’s development is crucial to fully benefitting from the opportunities AI provides. Below, we outline some of the key benefits and challenges of AI for AMLA:

Supporting national supervisory authorities in identifying troubling firms

AI can assist national supervisory authorities in identifying high-risk or non-compliant firms by analysing vast amounts of data that would not be possible relying on human analysts alone. Trend analysis tools, powered by AI, can identify suspicious patterns in financial behaviour and highlight areas of concern that would go unnoticed through traditional means.

“By learning from past financial crime patterns, AI can also predict future risks, equipping regulators and firms to act more proactively. Such predictive capabilities enable AMLA to be more responsive and efficient in its supervisory role, adapting its focus to emerging threats based on real-time data insights.”

Sylvie Matherat, Senior Global Advisor, Forvis Mazars in France

For AMLA, AI capabilities will allow it to highlight and address high-risk issues more quickly and efficiently, freeing up resources to focus on the most critical aspects of investigations. Predictive analytics, in particular, can help investigators understand emerging risks, allowing them to respond before problems escalate.

Coordinated training and awareness of AI tool capabilities and applications

AMLA will need to play an integral role in coordinating training on AI tools for both its internal staff and the staff of national supervisory authorities.

AMLA must ensure its employees, as well as the wider AML workforce, are educated and trained on the benefits and limitations of AI tools, understanding both the risks of over-reliance and the potential for human oversight.

In doing this, they will be well equipped to support AI-enabled innovation and explore the various use cases for AI in its supervisory work.

Moreover, AMLA will oversee how AI is being used in the financial services industry, fostering discussions with other regulators and firms about best practices. By leading these efforts, AMLA can ensure a comprehensive and informed approach to the use of AI in anti-financial crime frameworks.

Augmenting skills and capabilities of teams, not replacing them

Despite fears that AI will replace human jobs, its true strength lies in supporting and enhancing human capabilities. AI can automate labour-intensive tasks, such as sorting through vast amounts of transactional data, freeing up human teams to focus on higher-risk areas requiring judgment and expertise.

“AI tools are not substitutes for human decision-making but rather enablers that help investigators and compliance teams work more effectively.”

Luke Firmin, Head of Financial Crime, Forvis Mazars in the UK

Challenges and safeguards

The use of AI in AML operations presents unique challenges that must be carefully managed to reduce risk. Under the EU’s Artificial Intelligence Act, AI systems must adhere to strict requirements for transparency, accountability, and data protection. AMLA must ensure that AI tools comply with these regulations, balancing their implementation with the need to maintain trust in the financial system.

“Explainability and transparency are key issues in AI-driven AML efforts. Regulators and firms alike must be able to explain how AI systems arrive at their conclusions.”

Gregory Marchat, Group Head of Financial Services Advisory, Forvis Mazars in the UK

This is essential for fostering trust in AI decisions and ensuring they are understood by regulators, firms, and the public. Moreover, integrating AI systems with existing frameworks is crucial for maximising their benefits, ensuring that these advanced technologies are used to their full potential without creating new compliance risks.

Talent and skills development

A critical challenge facing AMLA and national authorities is finding and retaining staff with the necessary skills to work with AI. The deployment of AI in AML requires expertise that blends technology, data privacy, ethics, legal compliance, and financial crime prevention. Attracting professionals with this unique combination of skills is essential to the successful implementation of AI in AML operations.

AMLA will need to support ongoing training and development programs to ensure that its staff, and those within national authorities, remain up to date with the latest AI technologies and their implications. Additionally, member states will have a role to play in ensuring their national competent authorities (NCAs) have sufficient resources and expertise to meet these AI-related challenges.

AI in financial crime regulation: international perspectives

In addition to the EU’s approach to AI in anti-money laundering, other countries are actively integrating AI technologies to regulate financial crime. In the United States, AI is increasingly being used by financial regulators such as the Financial Crimes Enforcement Network (FinCEN) to monitor large datasets and detect suspicious financial activities in real-time. Machine learning models are also being used to identify fraudulent transactions and non-compliant actors at a scale that was previously unattainable.

In Singapore, the Monetary Authority of Singapore (MAS) has implemented AI-enabled solutions for risk-based assessments and real-time surveillance, detecting anomalies in financial transactions. This proactive use of AI enables regulators to identify money laundering risks before they escalate.

Similarly, in the United Kingdom, the Financial Conduct Authority (FCA) has been experimenting with AI for fraud detection and reducing false positives in transaction monitoring systems. By utilising AI-enabled technologies, the FCA seeks to streamline compliance processes while improving the accuracy of financial crime detection.

These global examples highlight the transformative role AI is playing in financial crime regulation, illustrating its potential to enhance the effectiveness and efficiency of supervisory frameworks worldwide.

“New technologies such as AI and machine learning offer tremendous opportunities for both banks and supervisors. However, to use these technologies safely and soundly, we need an adequate regulatory framework, proper supervisory oversight and an understanding by all users”.

Elizabeth McCaul, Member of the Supervisory Board of the European Central Bank in her speech on “The use of artificial intelligence to fight financial crime”, 13 July 2022.