Artificial intelligence can prevent money laundering

“Cognitive computing based on artificial intelligence can prevent money laundering much more effectively than hundreds of compliance officers. It’s only a matter of companies’ willingness to start using the new technology,” says Kasper W. Rost, CEO of the regtech company Calcabis.

By Carsten Jørgensen

​"A broad cooperation between financial institutions and the use of artificial intelligence can effectively stop criminals' attempts to launder their dirty money in banks," according to Kasper W. Rost, CEO of Calcabis.

The regtech company has developed a virtual expert platform (VEP) that transforms human intelligence into artificial intelligence (AI). Calcabis started in the medical industry, where they specifically interviewed medical experts and transferred their many years of accumulated knowledge into artificial intelligence.

"It took 2½ weeks to transfer the knowledge from a doctor with 30 years of experience to a computer. The AI platform is used the same way in the engineering industry, and obviously the banks also use cognitive computing based on artificial intelligence to handle their compliance work," says Kasper W. Rost.

Hiring of hundreds of new compliance officers doesn't necessarily make banks better at preventing money laundering. But if you take the most talented employees and transform their knowledge into artificial intelligence, it can be a very effective defence against criminal cash flows that are trying to be run through the bank.


Money launderers also use artificial intelligence

"If you're trying to supress money laundering and other criminal matters in banks, it's important to use both data-driven (machine learning, deep learning etc.) and rule-based (cognitive computing) AI. With data-driven AI, you can learn from monitoring and analysing data. For example, you can set up digital alarms for a wide range of transactions, and with cognitive computing you can then process and execute processes via AI that are usually manual," explains Kasper W. Rost, who then elaborates:

"It's important to remember that money launderers also use artificial intelligence to find the gaps in the defences. So banks have to be better at cooperating. If all banks and supervisors implemented AI in the right way and could agree on a general blockchain that links the banks' transactions and maybe also includes data from mobile phone companies, they would have a lot more parameters to analyse, thus increasing the likelihood of stopping a particular person or company from moving their laundered transactions to bank B when bank A has rejected them, and catch the criminals before they can even start their fraud."

He is surprised banks don't show a greater interest than they do in using the technology developed by Calcabis and other new regtech companies.

"It seems that many large banks are more focused on developing solutions themselves than purchasing ones that are already developed and work, which is a pity, because their strategy costs all of society huge amounts and in the end, they're best at making heavy, secure banking systems and not AI," says Kasper W. Rost.

He's confident in Calcabis because the company's compliance technology can be used in many industries. It will also be used more and more in the finance sector in the future, because Calcabis cooperates with one the major consulting houses that does compliance work for a number of major financial institutions.


Sektorens udvikling og vilkår; Fintech & IT