Vaccines are widely used as a medical preventative defence mechanism. They have greatly improved our odds of fighting disease for centuries. It is now possible to take the same concept and apply it to our fight against financial crime. Our financial system is systematically targeted by criminals that abuse its weaknesses. Financial crime behaves like a virus. It infects the subject (banks and other financial institutions) and often remains inert for a while. The host will begin to show symptoms of the illness but often only once the virus (financial crime) has copied itself several times and already caused irreversible damage.

How do vaccines work? 

There are various types of vaccines. Commonly a harmless synthetic substance, extracted from the original virus, is injected into the body of the subject. Once inside the body, it triggers an immune response that will cause antibodies to respond much as they would have on their first reaction to the actual virus [1] (figure 1).

How are financial crime vaccines created?

There is a clear analogy between medical vaccines and vaccines for financial services. We can now develop financial crime vaccines using synthetic data and simulations. Be it fraud or money laundering, financial crime techniques display transactional and demographic network characteristics which leave unique fingerprint patterns that can be modelled. Just like viruses there are different types of financial crime techniques, often referred to as typologies, that have certain characteristics (figure 2).

Synthetic data is becoming widely adopted by financial institutions. There are fewer data confidentiality issues and numerous benefits for AI applications [2] when using synthetic compared to sensitive real data. Information sharing is a key limiting factor in the fight against financial crime. Synthetic data allows firms to share knowledge about patterns of suspicious activity and effective counter methods without disclosing personal information or exposing themselves to regulatory breaches. Synthetically generated financial crime simulations now gives financial institutions the chance to build up their immunity to criminal techniques.

Under most circumstances, financial institutions perform rigorous stress testing and calibration of their control systems to create these antibodies. This can compromise confidential information and more often than not is too late to defend against the financial crimes. A vaccine using synthetic financial crime simulations allows financial institutions to build up their antibodies in a completely safe way within a fraction of the time and more confidently measure their level of protection.

By 2024 Gartner predicts that 60% of AI will be trained using synthetic data [3]. Is AI the solution for reducing the illicit profits earned from financial crime? There is ample evidence that points in that direction. With good quality data AI can certainly predict and detect most of the financial crime patterns that currently appear. It can even prevent patterns that are yet to be experienced by financial services. Last but not least, explainability plays a key crucial role in AI and synthetic data as it can justify to regulators why a transaction or a customer has, or has not, been identified as suspicious. 

Fighting the dynamic threat of financial crime

Financial crime is an evolving threat that requires continuous adjustments to the protective controls used to counter it. Similar to the mutation of viruses that require further booster vaccines, financial institutions require ongoing revisions and updates on the models they use for controlling financial abuse (figure 4). Synthetic data can be enriched periodically with new and emerging typologies that can threaten financial institutions. This is a key aspect in the fight against financial crime, since we are reducing the window of opportunity for a fraudster to abuse the financial system.

The rollout of financial crime vaccines

Collaboration is the answer to improve our odds in the fight against financial crime. A wider rollout of any financial crime vaccine should follow rigorous collaborative research and trials, as is demanded for medical vaccines [4] (figure 5). With proper interaction between regulators, academia and financial institutions collaborative models become possible, such as the Triple Helix AML proposed previously [5].

Having considered the similarities between financial crime and viruses one question now must be asked:

Is it time to roll out a vaccination programme to fight financial crime?

Get involved

Join the community of financial crime fighters helping to build, trial and rollout Financial Crime Vaccines by becoming a partner of our Financial Crime Vaccination programme.

References:

[1] How do vaccines work?. World Health Organisation (WHO). December 2020. https://www.who.int/news-room/feature-stories/detail/how-do-vaccines-work

[2] Lopez-Rojas, E.A., Axelsson, S., 2012. Money laundering detection using synthetic data, in: Annual Workshop of the Swedish Artificial Intelligence Society (SAIS). Linköping University Electronic Press, Linköpings universitet.

[3] Gartner, 2020. Predicts 2021: Data and Analytics Strategies to Govern, Scale and Transform Digital Business. Gartner.

[4] Kaufmann, Stefan & McElrath, M. & Lewis, David & Del Giudice, Giuseppe. (2014). Challenges and responses in human vaccine development. Current opinion in immunology. 28C. 18-26. 10.1016/j.coi.2014.01.009. 

[5] Lopez-Rojas, Edgar Alonso & Zoto, Erjon. (2019). Triple Helix Approach for Anti-Money Laundering (AML) Research Using Synthetic Data Generation Methods