Fighting Financial Crime with Artificial Intelligence

September 2021

Technology transformation is giving a face lift to every sector including financial services. However rapid digitization is a double-edged sword as criminals are becoming sophisticated and exploring security vulnerabilities and evading detection to move dirty money across the banking systems. It is estimated that up to USD $2 Trillion is laundered annually, which is the equivalent of almost 5% of the global GDP.

The consequences of financial crimes are devastating and have a wider economic, security and social impact. It allows perpetrators of illicit activities such as human and narcotics trafficking, and organized crime to continue and expand their operations, causing catastrophic losses to victims and wreaking havoc on societies. Activities such as terror financing poses a significant threat to national security – making this a serious threat for the world at large.

Financial institutions are collectively spending billions of dollars trying to prevent financial crime. Non-compliance to AML regulations leads to hefty fines by regulators. In 2020, financial institutions across the world paid USD $10.6 Billion in Anti-Money Laundering (AML) fines. Such incidents create commercial and reputational risks for the financial institutions. Despite all efforts, only a fraction of financial crime is identified.

Advanced Artificial Intelligence (AI) technologies can bring proactive detection, actionable insights and improved effectiveness and efficiencies to such programs.

AI for efficient operations and effective detection & investigations

Using AI, massive amount of structured and unstructured data can be analyzed intelligently, to reveal hidden connections and unveil actionable insights to identify, investigate and mitigate financial crime.

One of the most common applications of AI in financial crime programs has been in prioritizing high risk transactions flagged by traditional rules-based system and reducing false positives to allow investigators to spend valuable time on high priority cases.

While AI has been primarily used to gain operational efficiencies in financial crime programs, the real power of AI lies in the fact that it can combine operational efficiency with effective detection and investigation in KYC and AML programs. AI can harness billions of data points that are publicly available in the open web and combine with the organizations own data and subscription databases to reveal previously unattainable insights which aid in crime detection and investigation. In the absence of an effective Data sharing framework among Financial Institutions, open-source intelligence (OSINT) offers a vital edge to AML/CTF programs.

For example, traditional Know Your Customer (KYC) process is used by bank to identify if a potential customer is on a Sanctions, Politically Exposed Person (PEP) or a Special Interest Person (SIP) list. However, most people with criminal or extremism ties stay under the radar and are not captured in such watchlists. AI can tap into a vast pool open-source data to unearth such observations.

AI’s ability to contextualize and derive insights from vast amounts of unstructured data augments human investigations in areas such as Fraud or KYC, AML investigations. AI based models can replace archaic rules-based systems in identifying suspicious transactions more effectively.

As AI gains traction and becomes mainstream in combating financial crime, establishing trust in the outcome of AI and ML models through Explainable AI is a key factor in ensuring successful collaboration between financial institutes and regulators.

To stay one step ahead of threat actors, it is time financial institutions make AI a key element of their financial crime programs. It is imperative for FIs to have advanced detection and layered investigative capabilities by using AI solutions which can leverage a wide source of structured and unstructured data including legal and open-source data.

Share this post