Privacy Technology for Financial Intelligence: Dive into the future of financial intelligence with our cutting-edge exploration of privacy technologies.
Welcome to the New discovery Project. The mission of the project is to research technologies that can enhance collaboration for financial intelligence. Our project relates to advancing research and knowledge in financial technologies to combat financial crime particularly Anti-Money Laundering (AML). This project is brand new and being run for the first time in trimester 3.
Complex financial crimes are frequently orchestrated by well organised and experienced criminal groups. These illicit activities often encompass multiple business entities and individuals, with financial activities spanning various countries, effectively concealing the illegal operations through layers of subterfuge. Money laundering is one of the most severe financial crimes, exacting an annual economic toll on Australia estimated at 10 − 15 billion dollars.
To assist with preventing and understanding financial crime through utilising financial technology. WAs part of this project we will aim to explore prominent privacy technologies commonly researched and implemented in the field of financial intelligence. Some of these technologies include, but are not limited to, the following:
- Secure multiparty computation (SMC or MPC): A protocol that allows multiple parties to work together to achieve the common goal, whilst keeping secure of the information that are not meant to be shared. For example, private set intersection (PSI) only releases the common entities between two data bases.
- Homomorphic encryption (HE): An advanced cryptographic technique that allows certain arithmetic operations (e.g., addition and multiplication) to be performed on encrypted data (i.e., ciphertext) without needing the secret key. A common scenario of HE is to encrypt the sensitive data then outsource the encryption to a third party with computational or analytical advantages.
- Differential privacy (DP): A statistical method to release data base query outcomes with calibrated noise to guarantee no individual information can be reverse engineered from the outcomes. For example, carefully calibrated noise is added to the average income of a group of people to stop individual income to be inferred, should another query is allowed to exclude a certain individual from the data base. There are other popular privacy techniques that are often discussed and employed in financial intelligence, include hashing techniques, federated learning and so on
By the end of the trimester we would like to write a report exploring prominent privacy technologies (being Secure multiparty computation (SMC or MPC) Homomorphic Encryption, Differential Privacy among others).
- Initially we will look at a few common financial crime scenarios where intelligence sharing is inhibiting appropriate actions.
- Next the report will detail the technologies. It will describe what they are, how they work, pros and cons, current research on them. It will then discuss how they can assist with the scenarios. Finally it will outline further research areas