In 2016, the Reserve Bank of India (RBI) fined 13 Indian banks Rs 270 million for violating KYC regulations, while eight other banks were urged to put suitable safeguards in place and to be assessed regularly to ensure strict compliance with KYC requirements.
As of August 2019, the RBI had assessed Rs 265 million in fines on banks for failing to follow its account opening/operation instructions and end-use monitoring of funds. It has sparked a flurry of activities in the banking sector to improve compliance both in India and abroad.
Tracking Money Laundering
Most financial institutions rely on rules and procedures to learn about their customers and actions. However, money launderers have devised new methods to conceal their operations, which a typical rule-based system may be incapable of detecting. It leads to non-detection and non-reporting of suspicious transactions, resulting in non-compliance with Anti-Money Laundering/KYC laws.
There are three stages to any money laundering scheme: placement, layering, and integration. In the placement stage, illegally obtained funds are either invested in financial instruments or deposited into a bank account. Layering delivers money to third parties via wire transfers, checks, and orders. At this stage, the money is used to buy legal assets or to keep financing illegal businesses. Subsequently, funds earned dishonestly can be used openly.
As a result, banks must build a solid set of controls that will allow them to identify economic activity and transactions even when money launderers do their best to bypass the regulations. One possible approach is to utilise an Artificial Intelligence and Machine Language-powered AML transaction monitoring system.
Each of the three stages of money laundering are an opportunity to use AI techniques to look for signs of such activity. Machine learning techniques like support vector machines (SVMs) and random forests (RF) can be applied to huge, annotated bank datasets to categorise fraudulent transactions.
Banks typically employ data-driven strategies throughout the placement and stacking processes, as they can access transaction data. The final integration stage is hard to spot because the money has surpassed fraud-detection systems. At this point, AML could benefit from applying cutting-edge AI techniques, such as entity relationship extraction from massive social media and news data.
AI’s Fast Acceptance
With many tech companies racing to integrate AI technology into a variety of businesses and industries in the last few months, Google Cloud has just introduced an AI anti-money-laundering tool for banks.
But practical applications of AI in AML programmes have been around for some time now:
* Danske Bank used an AI-powered system to detect fraudulent transactions in 2018. The system continuously analysed customer data and transaction patterns to learn and identify risks. As a result, Danske Bank enhanced its AML programme's accuracy by 60 percent, lowering false positives.
* In 2019, HSBC introduced an AI-powered solution to automate anti-money laundering (AML) processes. Machine learning algorithms were employed to analyse consumer data and identify suspicious transactions. It enabled HSBC to minimise the time necessary for AML reviews, enhance programme accuracy, and save $400,000 annually.
* Standard Chartered used an AI-powered technology to automate their AML operations in 2020. The system analysed real-time customer data and transaction patterns to identify risks. As a result, Standard Chartered cut compliance review time by 40 percent while simultaneously boosting the accuracy of its AML programme.
* JPMorgan Chase implemented an AI-powered technology to boost their AML programme in 2021. Machine learning algorithms were employed to analyse client data and identify potential dangers. JPMorgan Chase minimised false positives by 95 percent while enhancing its AML programme's accuracy.
How AI Makes A Difference
There is no doubt that AI will come to be used by banks on several frontiers to crack down on money laundering.
- Transaction Monitoring: AI can look at a massive amount of transaction data in real-time, which lets it find suspicious activities quickly that need to be looked into further.
- Risk Assessment: AI can be used to find high-risk customers and transactions so that AML workers can focus their efforts where they are needed most.
- Customer Due Diligence (CDD): AI can automate the CDD process by verifying customer names, checking against sanction lists, and finding possible fraud risks.
- Suspicious Activity Reporting (SAR): AI can help find unusual transaction trends and behaviours that could be signs of money laundering.
- Enhanced Screening: AI can improve the accuracy of screening by finding possible false positives and lowering the number of manual reviews that need to be done.
- Predictive analytics: Based on past data and behaviour trends, AI can use predictive analytics to predict possible money laundering.
Compliance experts claim AI-driven anti-money-laundering systems have improved over the years. But not everyone in the banking sector are convinced of their ability to replace a human's ability to determine where the money laundering risks genuinely lie. But with banking regulators using AI tools to uncover AML failures of banks and imposing hefty penalties, banks have no choice but to hop quickly on the AI bandwagon.
Neha Jogi is a freelance technology writer. Views expressed are personal, and do not represent the stand of the publication.
Discover the latest Business News, Sensex, and Nifty updates. Obtain Personal Finance insights, tax queries, and expert opinions on Moneycontrol or download the Moneycontrol App to stay updated!
