Artificial Intelligence at HSBC – Money laundering use Case

HSBC Bank has worked with several AI vendors and provided evidence of success that other banks can be capable to study and take benefit of.

HSBC’s AI-enabled Anti Money Laundering Solution. In 2018, HSBC partnered with Ayasdi to develop an AI-enabled anti-money laundering solution. The software can allegedly identify patterns in historical data that might point toward money laundering, which helps the bank stop payments before they violate regulations. Ayasdi claims to have reduced HSBC’s false positives by 20% and found many behavioral models directly related to fraud. The solution was developed in collaboration with HSBC’s IT team and Ayasdi’s data scientists and developers. 

The IT team allegedly helped Ayasdi access and parse the bank’s AML data. The bank could then understand these models because they were made using terms they were already familiar with. Ayasdi’s solutions are mainly based on anomaly detection technology, which is useful for recognizing deviations from a pre-established norm. They claim that their software analyzes the sources and destinations of client payments to make certain the funds are coming from legitimate sources. Anomaly detection software appears to have worked well for HSBC along with other banks are looking to improve their defense against money laundering. It is because well-trained algorithms might recognize deviations much quicker than human analysts at computers. 

HSBC’s Interest in AI for trade flow and document research. In 2019, HSBC announced a partnership with Element. AI, a firm that mainly offers AI solutions for business flows and document searches. They also offer AI business management solutions like an access governor that determines which employees can access which sets of data. The solution will help them predict which products and service solutions their clients might need in the future. HSBC can be capable to achieve its goal of global compliance using AI resources from Element. An NLP’s solution will allow them to automatically tag new info with metadata for better research and transparency. 

This allows a client company to more precisely retrieve the data requested by a client or auditor. Element AI’s solutions more than likely run on some mixture of natural language processing and predictive analytics technology. NLP might be utilized to create documents search applications and automatically tag documents with metadata. HSBC probably intends to use Element’s predictive analytics engine to study client data for info that could indicate possible problems. For instance, Element’s software can be capable to detect metadata tags pertaining to the kind of client service issue for each email. They might then determine that most of their client service tickets come from their mobile application not recognizing a confirmation code. Almost all these clients can report a problem where the application wouldn’t allow them to log in until they’ve requested a second code and used it for verification. HSBC can be capable to recognize this problem as it arises and start working on a fix earlier than if detected and reported manually by client support agents.