Artificial Intelligence is reaffirming itself as one of the key resources for the financial sector, which is increasingly challenged by issues such as anti-money laundering and counter-terrorism financing.
For several years now, the FATF (Financial Action Task Force), the global regulatory body that sets international standards to prevent criminal activities related to money laundering and terrorism financing, has been encouraging the use of new technologies to support financial institutions in assessing AML/CFT (Anti-Money Laundering/Counter Financing of Terrorism) risks. This aims to ensure greater accuracy, timeliness, and completeness while strengthening the effectiveness of regulatory and preventive measures.
Within this context, the research project FinVA - Financial Intelligent Virtual Assistant was launched by the Institute of Information Systems and Networking (ISIN) in collaboration with business partners CubeFinance and QBT. Funded by the Swiss Innovation Agency (Innosuisse), the project led to the development of a cutting-edge AI-based solution that ensures the security of sensitive data while efficiently identifying high-risk relationships.
The platform is designed to revolutionize how financial institutions approach Know Your Customer (KYC), continuous monitoring, and the organization of customer identification and verification processes.
A distinctive feature of FinVA is its ability to "generalize." Thanks to advanced machine learning algorithms and the uniqueness of the expert knowledge formalization methodology, the platform is able to manage Customer Due Diligence (CDD) processes, a standardized procedure for customer verification and the assessment of associated risk. It assigns a risk score to contractual relationships and analyzes transactions in the context of the customer where they take place.
These features not only enable an accurate predictive analysis of contractual risk, allowing for risk anticipation and improved compliance measures effectiveness. The most significant advantage is that they also allow for an 85% reduction in so-called “false positives”—transactions that traditional AML analysis systems flag as risky and potentially reportable to Supervisory Authorities, causing significant impacts in terms of time, costs, and resources.
In conclusion, FinVA represents a major step forward for financial institutions, demonstrating how Artificial Intelligence can be leveraged not only to enhance compliance processes but also to reduce costs, improve security, and increase effectiveness in the fight against illicit financial activities.