Head to head
38
FEBRUARY 2022
AI in financial services Artificial intelligence stepped out of science fiction years ago and is at work in Luxembourg in insurance, financial services and fraud.
Andreas Braun, director, artificial Intelligence and data science, PwC Luxembourg
How is artificial intelligence used in AML has flowed into other anomchanging the way financial services aly detection tools, such as cybersecurity are delivered in Luxembourg? for financial institutions. ANDREAS BRAUN It’s a broad question AJAY BALI I would say that AI has already because AI has changed the way finan- changed the way we deliver financial sercial services are offered on multiple vices. With something like AML, the volfronts. Let’s take the retail customer. A ume of data is huge. What ML can do is machine learning-based analysis of your pull together different combinations of income and expenditure is used by the data and flag it to the decision makers. bank. For example, if you are increasing Often the customer is unaware that AI is your spending they can sell a credit prod- being applied in very different ways to uct. If you are saving it is an opportunity make decisions on their behalf. to offer plans for retirement. BRAUN In Luxembourg, AI also helps a In the back office, where the cost of lot with the multilingual factor. The transcompliance is ever increasing, we see lation tools that AI can deliver really help automation powered by ML and natural Luxembourg financial services compalanguage processing to look at documents, nies with exposure to markets like Asia. pick out the important parts and make the life of the back office easier. In ‘know You say that the customer is unaware. your customer’ processes, it has been Are there best practices for the use common for years to take into account of AI in financial services? a customer’s social networks and to carry BALI There are governance and best pracout a sentiment analysis. In the past, this tices around how data is fed into AI in would have been done by Googling, now order to prevent bias--for example, when it’s done through sophisticated ML. The getting a loan approved. If loans are fresame principle applies for anti-money quently granted to customers in one postlaundering, where ML can pick out anom- code and not to customers from another, alous behaviour. In fact, what has been governance is needed in the way you feed data to prevent AI from automatically implementing a bias. Take facial recognition for the KYC process. Unless you train AI with multiple images of people with different appearances, you could build in a lack of recognition and therefore a bias. BRAUN The European Commission is defining best practices in AI to allow the ongoing functioning of markets but also to ensure that critical things affecting the consumer are being addressed. This includes bias. BALI There are a number of risks with AI. Not only bias but also a business risk, in that the data you’re being fed may have no integrity, a financial risk if AI forms the basis of financial decisions and privacy risk.
“ The world we are living in is getting more and more complex”