Harnessing AI in Anti-Money Laundering: A Look at Google Cloud’s AML AI Solution

Traditional AML monitoring tools typically rely on predefined rules based on an institution’s risk appetite and practices. When transactions match these predefined rules, they are flagged as high-risk and subject to manual review by subject matter experts. This type of monitoring system, while having served its purpose in the past, presents some significant challenges. 

The primary issue with rule-based AML engines is the high occurrence of “false positives”, where transactions are incorrectly flagged as suspicious. More than 95% of AML transaction alerts are classified as false positives. These false alarms require manual reviews, which increase operational costs and workload. Additionally, there is always the risk of genuine fraudulent transactions slipping through the cracks, leading to potential regulatory breaches and financial losses.

Google Cloud’s AI-Based AML Solution

In contrast to traditional transaction monitoring tools, Google Cloud’s AML AI tool uses machine learning to generate customer financial crime (FinCrime) risk scores. Google said it accomplishes this by analysing a financial institution’s other systems and data, such as transactional patterns, network behaviour, and Know Your Customer (KYC) data.

Prominent financial institutions such as HSBC, Banco Bradesco, and digital bank Lunar already leverage this innovative solution. HSBC’s Group Head of Financial Crime Risk and Compliance said Google’s AML AI tool already shows the prospects of changing how institutions fight financial crime:

“Google’s models are already demonstrating the tremendous potential of machine learning to transform anti-financial crime efforts in the industry at large.”

Jennifer Calvery, Group Head of Financial Crime Risk and Compliance, HSBC

AI’s Edge in Financial Crime Risk Management

In a 2021 report, the Financial Action Task Force (FATF), an intergovernmental organisation, highlighted the potential of new technologies in combating money laundering and terrorism financing. The report suggested that machine learning, a subset of AI, significantly enhances these efforts as it learns from other systems, reduces reliance on manual contributions, minimises false positives, and aids in detecting complex suspicious transactions.

The FATF identified four key areas where machine learning could make a marked difference:

• Customer identification and verification (ID&V)

• Monitoring of business relationships and behavioural and transactional analysis

• Identification and implementation of regulatory changes

• Automated data reporting

Important Considerations

Despite the promising potential of Google’s and other companies’ AI-based AML solutions, financial institutions must fully understand the cost implications and consider data privacy, model risk, and other risks that may arise or change when adopting the tool.

Careful cost forecasting is needed before adopting cloud-based services. According to Gartner, cost overruns are common, with 60% of infrastructure and operations leaders predicted to encounter public cloud cost overruns through 2024. In the case of the Google AML AI tool, one of its two cost components is “the number of banking customers the service is used for.” That implies that the cost would be the same for a bank with a higher risk profile of customers as another with a lower risk profile if they both have the same number of customers.

Additionally, institutions, particularly those adhering to EU GDPR and jurisdictions with similar regulations like the Abu Dhabi Global Market (ADGM), must be cautious to ensure data privacy regulations are adhered to before using a cloud service provider.

How Várri Consultancy Can Assist

Navigating the landscape of AI-driven AML solutions can be challenging, especially considering the complexities of cost implications, data privacy risks, and compliance with international regulations.

At Várri Consultancy, we recognise that an effective AML solution isn’t just about preventing financial crime – it’s about enhancing your overall business performance. We, therefore, take a holistic approach to understanding how these solutions impact all aspects of your organisation, including customer experience and operational efficiency, and how it may impact other risks your organisation faces. Together with our collaboration partners, we’re equipped to guide organisations through the assessment of their AML requirements, the selection of suitable solutions, and the development of an effective policy:

1. Comprehensive AML Assessment: We thoroughly evaluate your current AML strategies and practices, identifying gaps and areas of risk. We also consider the potential impact of new AML solutions on your customer service, operational efficiency, and overall business performance.

2. Solution Selection: We guide you in selecting the right AML solutions that best fit your needs. This includes evaluating potential AI tools, considering not just their functionality and cost but also their potential to enhance customer experience and other critical business processes.

3. Policy Framework Development: To ensure the effective use of your chosen AI tool, we assist in developing a robust policy framework. This framework will focus on risk management and compliance and consider how implementing the AML solution can harmonise with other business functions, fostering a more integrated and efficient operation.

4. Implementation Support and Training: We offer extensive support in deploying your new AML solution, providing necessary training to your staff. We ensure that they understand how the new system can improve not just compliance and risk management, but also customer interactions and overall service delivery.

5. Continuous Monitoring and Support: We offer ongoing support, monitoring the effectiveness of your AML solution and suggesting adjustments as needed. Our comprehensive approach means we’re not just looking at compliance but evaluating how the new system impacts customer experience, operational efficiency, and other essential metrics.

Contact us today to learn more about how Várri Consultancy can help your organisation navigate the complexities of financial crime and AML. For insights into data privacy risks in cross-border data transfers, read our article on “Recent Developments in EU-U.S. Data Transfer Regulations.”