Navigating FRAML: The Fusion of Fraud and AML

Roy Prayikulam, SVP of Risk & Fraud at INFORM, reveals best practices in the fight against financial crime, including Fraud Detection and Anti-Money Laundering (FRAML) tooling.

In an era where digital transactions are a daily norm, the shadows of financial crime grow more extensive, demanding a revolution in protective measures. To combat this sophisticated threat landscape, the financial sector's crime fighters are evolving, too, uniting Fraud Detection and Anti-Money Laundering (FRAML). But how prepared are we? And how realistic is it that participants in the payment cycle will really switch to FRAML as a standard process?

‘Empower our Fight Against Financial Crime’ - Under this motto, no less than 137 representatives from 24 banks and payment service providers, eleven countries, and four continents gathered at the end of September 2023 to exchange information on the current state-of-the-art and best practices in the fight against financial crime at a network meeting organised by INFORM. ‘FRAML’ was among the focus topics discussed.

In recent times, the global financial realm has found itself enmeshed in a complex web of innovative yet nefarious financial crimes, amounting to over USD 800 billion laundered annually. The emergence of FRAML symbolises a proactive stride towards fostering a synergised front, joining Fraud Detection and Anti-Money Laundering mechanisms to navigate the complex landscape of modern-day financial crimes. But is it hype or reality?

Traditionally, Fraud Detection and Anti-Money Laundering (AML) operated as distinct entities, each with its own methodologies, data systems, and regulatory compliance mechanisms. Each had its own set of protocols and technologies, with minimal intercommunication. This segregation fostered inefficiencies, as both departments aimed to combat financial crimes, albeit from different angles. The siloed approach often led to redundant investigations and overlooked threats, as demonstrated by a project at an Asian bank, revealing a half-million-dollar risk exposure in potential charge-offs when viewed holistically rather than in isolation. Thus, the historical separation has been under scrutiny, with industry debates and reports advocating for a more integrated approach to optimise resources and enhance financial crime detection.

Regulatory bodies are progressively accentuating the adoption of innovative technologies to fortify the effectiveness of fraud and financial crime risk management. The US-American Anti-Money Laundering Act of 2020, for instance, underscores the necessity for financial institutions to adopt risk-based programs to curb money laundering and terrorist financing, exemplifying the regulatory impetus towards a unified, technologically advanced approach to combating financial crimes.

FRAML, especially when driven by AI, is believed to aid in more accurate results by automating data sharing and insights between functions that traditionally operated in silos. For instance, in the digital-first banking era, the hyper-digitisation of fraud, identified by the Treasury as a major predicate offense for money laundering, has pushed the need for an integrated FRAML compliance framework. This shift enables real-time detection and analysis of illicit activities as well as predictive modelling, significantly reducing false positives.

During the above-mentioned exchange among experts, money mule scenarios as well as Continuous Due Diligence (CDD) requirements to protect merchants were mentioned as frequent use cases for FRAML. INFORM has also identified another use case and developed a software solution for telecommunication service providers in mobile money, where customers use their phones as a ‘bank account’ for financial operations, like registrations, logins, financial transactions, or loan requests. In this field, best practice AI-based solutions on one end-to-end platform integrate fraud prevention, AML compliance, and credit management throughout the customer journey. For example, when a new customer signs up for registration, the AI performs customer segmentation and risk scoring based on various input data, but at the same time scans the customer against watchlists and sanction lists. Comparable integration of fraud-related and compliance-related evaluations has been implemented for the entire customer lifecycle and happens in real time dynamically triggered by each new action a customer takes.

While FRAML presents undeniable advantages, it's not without its regulatory hurdles. Financial institutions must navigate a labyrinth of compliance requirements, some of which can inadvertently impede integrated approaches. This is one of the main reasons why processes of this kind have not yet become a universal standard in the banking sector. But seamless integration of fraud detection and AML is not only an organisational issue, albeit also a technological one. And at least in this area, there are already clear answers to the question of which technology effectively promotes FRAML.

At the heart of this evolution – again – stands AI, particularly an approach called Hybrid AI. This technology combines the decision-making process of classical knowledge-driven AI techniques (e.g., Fuzzy Logic or dynamic profiling) with the adaptive, learning capabilities of data-driven techniques (e.g., Machine Learning). In the context of FRAML, this means that systems are not just rule-followers; they can now become rule-makers. They analyse vast datasets, learn from them, and identify hidden patterns of fraudulent activity and money laundering that would escape human analysts or conventional rules-based systems.

For instance, in detecting complex financial crimes, Hybrid AI systems can flag unusual transaction behaviours while considering contextual information, reducing false positives without compromising detection rates. They adapt to new fraud tactics and money laundering schemes. These technologies foster adaptability, enabling systems to evolve with emerging fraud patterns. At the same time, the human remains in the loop and can still further modify the dynamically adapting rules based on their knowledge, simulate new rules and, above all, immediately take them into live operation.

While the many benefits of FRAML, including cost efficiency, a broader perspective on financial crime, and of course the increased precision of risk assessment, are obvious, the operational alignment of fraud management teams and anti-money laundering teams, as well as complying with data-sharing regulations, remain a challenge.

When successfully deployed and powered by an AI-solution, whether in isolated use cases in the banking sector or holistically in the telecommunications industry, FRAML's integrated approach not only strengthens defences against sophisticated financial threats but also promotes operational efficiency.

About Roy Prayikulam

Roy Prayikulam is SVP Risk & Fraud at INFORM. After graduating from RWTH Aachen University, he started his career in financial crime prevention. Prayikulam has extensive experience in IT integration projects for the financial sector, for example in the areas of acquiring, card issuing, internet banking, and AML compliance.

About INFORM

INFORM develops software to optimise business processes using artificial intelligence and advanced mathematics of operations research. Founded in 1969, the company promotes sustainable value creation in various industries through optimised decision-making. Its solutions are tailored to specific industry requirements and help customers worldwide to operate more resiliently and sustainably with greater success, including industries like financial institutions and telecommunications.