This newsletter summarises key themes and challenges discussed in recent supervisory exchanges and workshops related to Anti-Money Laundering and Countering the Financing of Terrorism (AML/CFT). These insights, drawn from supervisory stocktakes and an industry workshop, aim to provide useful perspectives for supervisors and financial institutions in their ongoing efforts to enhance risk management practices. The Committee believes the information provided may be useful for both supervisors and banks in their day-to-day activities.
- The Basel Committee conducted a comprehensive stocktake of AML/CFT data and methodologies across 19 jurisdictions to identify key supervisory practices and challenges in assessing money laundering and terrorist financing (ML/TF) risks.
- This review identifies diverse approaches that enhance efficiency in risk-based supervision and effective supervisory judgment, including: (i) data collection and use; (ii) risk assessment methodologies; and (iii) key challenges.
- Recent industry outreach included information sharing on transaction monitoring systems (TMS) and the growing role of Artificial Intelligence (AI), emphasising the critical role of data, technology and governance in combating financial crime.
- The Basel Committee remains committed to supporting supervisors and fostering dialogue between supervisors and the industry to advance the effectiveness of AML/CFT supervision.
Enhancing AML/CFT Risk Assessments: key findings from a supervisory stocktake
Accurate risk assessment at supervised entities is essential for effective risk-based supervision and efficient resource allocation. The Basel Committee reviewed practices across 19 jurisdictions, offering insights to enhance AML/CFT data collection and risk assessment methodologies. The stocktake revealed that a sound understanding of money laundering and terrorist financing (ML/TF) risks is critical for effective, risk-based supervision. Authorities rely on diverse data sources, including AML/CFT questionnaires, Financial Intelligence Units (FIUs) and audit reports, to collect information about customer profiles, transaction types, geographic risks and the quality of control frameworks. These data points are essential for identifying and managing risks at the entity level.
AML/CFT questionnaires emerged as a cornerstone of data collection, with most authorities gathering over 100 data points annually to inform their risk assessments. However, challenges remain in efficiently processing and analysing this data, as manual methods are still widely used. Supervisory technology (suptech) is being explored as a potential solution to enhance automation and improve the quality of insights derived from the collected data.
Risk assessments typically involve balancing inherent risks, such as customer profiles and geographic exposures, with the adequacy of AML/CFT controls. Authorities often use a combination of formula-based and matrix-based approaches to calculate residual risks, which guide supervisory attention to high-risk areas. The review also highlighted the importance of regularly updating methodologies to reflect the dynamic nature of ML/TF risks and the evolving regulatory landscape. However, challenges persist, including ensuring data quality and completeness, balancing inherent risks with control measures while maintaining a focus on high-risk areas, managing cross-border complexities and addressing the risks posed by diverse regulatory expectations.
The role of AI in transaction monitoring: insights from an industry workshop
In November 2024, the Basel Committee hosted a workshop to explore the growing role of Artificial Intelligence (AI) and Machine Learning (ML) in transaction monitoring systems (TMS).
The workshop highlighted the potential of AI and ML to enhance transaction monitoring by improving anomaly detection, customer segmentation and dynamic threshold management. These technologies help financial institutions process and integrate large volumes of data, enabling a more comprehensive understanding of customer activity and the detection of suspicious patterns. However, the benefits of AI depend heavily on access to high-quality data, which requires robust management frameworks and collaboration between public and private sectors.
The discussions underscored the importance of human oversight. Participants noted that AI systems must complement traditional monitoring approaches rather than replace them. Human involvement remains critical for validating models, reviewing alerts and making final decisions on suspicious transaction reports (STRs). Moreover, the explainability of AI-generated results is essential to ensure compliance with regulatory expectations and to build trust among stakeholders.
The workshop also addressed the challenges of implementing AI in AML/CFT efforts. Poor-quality data, biases in algorithms and the resource-intensive nature of AI adoption were identified by participants as key obstacles. Participants noted that institutions must invest in skilled personnel, infrastructure and continuous monitoring to ensure the effectiveness of these advanced systems. Some participants indicated that regulatory uncertainty around AI use in AML/CFT further complicates adoption, highlighting the utility of clear international guidance and public-private cooperation.
The insights from the supervisory stocktake and the AMLEG workshop emphasise the critical role of data, technology and governance in combating financial crime. Collaboration across sectors and jurisdictions will be essential to foster the development of effective AML/CFT practices, including enhanced data sharing. The Basel Committee remains committed to supporting these efforts through ongoing dialogue and the sharing of best practices.