Imagine a bank customer who lives in the UK. This morning, she’s made handful of purchases in her neighbourhood – a coffee at her favourite cafe, some groceries from her local shop, some fuel at a local petrol station. Then, all of a sudden, she spends almost £1000 on electronics at a store in Malaysia. Typically, her bank would flag this unusual purchase – yet the money may have already been taken from the account and would be near impossible to claw back.

To address this kind of fraud, a growing number of banks, payment providers, credit card companies and other Financial Services (FS) businesses are turning to artificial intelligence (AI). According to a recent survey by NVIDIA, a chipmaker, fraud detection is the top reason that financial services businesses are investing in AI. Indeed, some 31% say this is their primary reason for deploying the tech.

Microsoft is one of the leading providers of AI programmes that are designed to tackle financial fraud. Using Azure, Microsoft’s enormous cloud computing platform, they train AI programmes on vast amounts of data which then helps FS firms identify and prevent fraud.

Related: What is Microsoft Cloud for Financial Services… and do you need it?

What is AI for financial services?

Artificial intelligence refers to a variety of competing technologies that can imitate human cognitive capabilities, analyse information, and spot patterns. With cloud computing environments such as Microsoft Azure, AI programmes benefit from vast amounts of computing power, as well as bigger data sets to learn from – thereby making them more useful and accurate.

How can artificial intelligence tackle banking fraud?

AI programmes can be trained to identify a variety of types of fraud. Some of the most common uses for AI in addressing financial services fraud include:

Personalised customer behaviour profiles

One of the most compelling use cases for AI in financial services is its ability to learn about each individual customer and their normal behaviour. As with the example at the beginning of this article, an AI programme could be trained to learn about each of your customers’ typical spending patterns, then immediately alert you when an unusual purchase occurs.

Now, many banks already have automatic rules to prevent fraud. A common approach is to freeze an account if spending occurs in countries that are deemed risky. The problem, however, is that this is hardly subtle. For example, an expat living in the UK might return to his home country and immediately have his card frozen for no good reason.

But with AI, fraud detection can be much more sensitive. In this example, the AI might ‘learn’ that this expat regularly sends money to bank accounts in their home country and has visited before. It therefore would not instantly freeze his card when he goes home to visit.

Tackling chargebacks

A chargeback is a form of E-commerce fraud whereby a customer instructs their bank to claim a refund for a product purchased online, even though they received the goods. In chargeback fraud, customers falsely claim they were charged for something they didn’t buy, thereby getting their money back (to the detriment of the vendor).

Banks rarely have the time to investigate if the customer is behaving fraudulently. However, by using an AI programme, a bank may find it easier to spot patterns in chargeback behaviour, such as an individual regularly making these requests.

Creating more accurate fraud scores

Another way that AI can help banks to identify and tackle fraud is by constantly learning about purchasing behaviour and building ever more accurate fraud scores. Machine learning algorithms can monitor variables such as transaction volume, time, IP address, frequency of card use, location and other information to identify fraudulent behaviour as soon as it happens.

eBook: The cloud era of financial services

Benefits of using AI to tackle financial services fraud

Investing in artificial intelligence tools to tackle financial services fraud provides a number of benefits:

  • Allows you to save huge amounts of time analysing data manually
  • Means you can reassign human resources to tackling new or complex forms of fraud
  • Allows you to quickly collect evidence for fraud investigations
  • Supports know your customer (KYC) and anti-money laundering (AML) processes
  • AI can save banks significant amounts of time and money, and improve customer relationships

How do Microsoft’s anti-fraud AI tools work?

When banks and other FS businesses use the Microsoft Cloud for Financial Services, they get immediate access to a number of AI algorithms which can be applied to their own data right away. This includes programmes to build customer profiles, identify fraudulent merchant activity, build risk profiles and more.

Microsoft’s AI tools are already being used by several major financial services businesses to tackle fraud, including Mastercard, Chargebacks911 and Abu Dhabi Islamic Bank, among others.

Use the best tools to tackle fraud

At FITTS, we help financial services firms to deploy Microsoft technology that is specifically designed for banks, insurers, payment service providers, fintech start-ups and funds. We can support you to implement Microsoft’s AI tools to tackle fraud, as well as enhance customer relationships, onboard faster, and improve internal collaboration and communication.

To learn more, contact us today.