After completing his PhD in Computer Science at the University of Nevada and spending the next 8 years working for research labs and companies such as Berkeley National Lab and Nokia, Dr. Leandro Loss moved to Philadelphia to join QuantaVerse, a startup focused in applying data science to financial crime detection. There, he has been serving as their Lead Data Scientist designing and implementing AI solutions to Anti-Money Laundering systems. His research and work experience have been primarily focused in machine learning and artificial intelligence, which he tries to always apply to projects with positive social impact. In parallel to his work in the industry, Dr. Loss is an adjunct professor at the International Technological University, in San Jose, California, and a visiting professor at ESSCA School of Management, in Shanghai, China.
The Role of Machine Learning and Artificial Intelligence in the Fight Against Money Laundering and Terrorist Financing
Money laundering (ML) and terrorist financing (TF) account for roughly US $1-2 trillion annually, or 2 to 5% of global GDP. Yet it is estimated that less than 1% of global illicit financial flows are currently detected and seized by authorities. On the front line in the fight against ML and TF are human Anti-Money Laundering (AML) investigators tasked with identifying and reporting suspicious activities, by reviewing thousands of transactions flagged daily by current rule-based transaction monitoring systems to determine if they are, indeed, suspicious. Even worse than the financial damage caused to financial institutions, the proceedings from ML are collected from every day crimes such as drug dealing, bribery and sex trafficking, while TF results in a heavy toll on human lives. Despite the value that Machine Learning and Artificial Intelligence have added to many businesses and domains, financial institutions (the main players in ML and TF schemes) have been virtually unchanged by them. For the last 20 years, typical financial crime investigators are equipped with spreadsheet programs and internet browsers in which they use to track thousands of transactions a day and cross-check names, addresses, purpose of business, entity relationships, transaction rationale, etc, in literally tens of government databases and websites. Their goal is to detect anomalies which are analyzed and further classified as suspicious or normal. In this talk, we will describe how Machine Learning and Artificial Intelligence can (and should) be used by financial institutions to address ML and TF. We will present a general purpose framework that is capable of (1) automating the collection of evidence, (2) detecting anomalies, and (3) ultimately adjudicating intent. Finally, we will discuss how human investigators and machines can jointly work to detect and prevent financial crimes in the globalized financial scenario, along with the new roles investigators will take in training and supervising the performance of intelligent algorithms.