At the recent week-long FIBA AML Virtual Conference 2020, the application of innovative technology for compliance (especially AML and CFT) was a recurring theme discussed by expert panelists representing banks, consulting firms, and solution providers.

A tech-focused session on the third day of the event, titled “Managing Financial Crime Risk through the Use of AI and Machine Learning,” featured a number of panelists who shared perspectives and use cases illustrating how banks are leveraging new technologies to reduce false positives without compromising effectiveness.

The panel’s moderator began by noting that the most common strategy banks and other financial institutions use for transaction monitoring tasks is home-grown or vendor-provided rules-based TMS. While this was the right approach historically, he argued that those dedicated to illicit activities work around the legacy technology’s limitations. One panelist from a large financial institution agreed and stated that static rules built into TMS create copious false positives resulting in staff spending countless hours clearing alerts and documenting why these alerts are not suspicious. The panel discussion turned to how proven data analysis techniques and technologies, such as AI and machine learning, are providing relief by reducing false positives, automating investigations, and even uncovering false negatives.

Phil McLaughlin, Chief Information Officer of QuantaVerse, discussed the benefits of utilizing keyword data (from RPA solutions and other sources) to run sentiment analysis using natural language processing (NLP) algorithms. QuantaVerse uses this technique to automatically determine sentiment associated with adverse media of specific alerted entities. Along with automatically ferreting out other essential information, this technology can, for example, differentiate if named entities are innocents in an article or the bad actors.

Phil also raised up the example of economic purpose as a critical determiner of risk. NLP techniques and trained AI models can intelligently determine if transacting entities are engaged in complementary lines of business and if transactions follow general business logic. For example, a steel manufacturer paying for iron ore makes perfect economic sense while the same manufacturer purchasing tons of bananas from a fruit exporter would raise a flag. While a human investigator would clearly understand this, it would take valuable time to pull together. The QuantaVerse AI platform automates 70% of investigative tasks, enabling AML teams to focus their time and talents on the most complex cases.

On the final day of the FIBA conference, QuantaVerse Founder and CEO, David McLaughlin, participated on the “Customer Profiling, Use of Innovative Technologies in Onboarding and Risk Assessment” panel. David Schwartz, President and CEO of FIBA, set up the panel discussion by emphasizing the importance and impact that innovative technologies have on risk assessment and the customer onboarding process.

David McLaughlin spoke to the advantages of accurately automating the entity resolution work associated with transacting customers and counterparties. An incomplete picture of risk can be caused by poor data entry such as abbreviated or missing names and addresses. David explained that web data parsing, named entity extraction, and sentiment analysis can be employed to automate the process of finding reputational risks associated with an entity in question. With proper entity resolution, data can be utilized to group together transacting parties with link analysis. And through network visualization, investigative staff can better see and interpret transactional connections in a network, enabling them to gain insights into the scope of activities of a risky group of customers.

To schedule a meeting with QuantaVerse and hear more about AI-powered solutions available today to solve for false positives, adverse media/sentiment analysis, entity resolution, and more, please visit: