AI including ‘magic to fraud detection’ Mastercard boss says

  • Mastercard is using generative AI to get one step ahead of criminals 
  • The payments network giant says it can help to identify compromised cards

When you think of generative AI and fraud, your mind might not immediately conjure up images of it being used as a force for good.

Quite the opposite – criminals have been using new advances in AI with great success to trick their victims out of money. It allows them to create fake ads and scam websites faster than ever before.

Recently This is Money revealed how criminals are using AI to make realistic counterfeit tickets to events from football matches to concerts

But card network giants – and fierce rivals – Mastercard and Visa have both recently revealed how they are using generative AI to beat criminals at their own game.

Mastercard and Visa are beating criminals at their own game by using AI to predict stolen card details 

Mastercard says it has added AI to its armoury for predictive powers so it can rapidly detect when criminals are using stolen debit and credit card details. 

The use of AI in fraud detection is nothing new, as banks and the payments industry have been using it in different forms for the last 20 years. 

But using generative AI in fraud detection is breaking new ground, according to Rohit Chauhan, vice president of AI at Mastercard.

He told This is Money that ‘generative AI is adding some magic to fraud detection’.

Every year, fraudsters steal millions of credit and debit card numbers through spyware, malware, and other clandestine practices such as ‘card skimming’.

Card skimming is when card details are compromised when shopping online, either by malicious code on retailers’ websites or by tricking consumers with false advertising on social media sites.

Once criminals have a hold of your card details, they can then use them to go on a shopping spree and even sell your card data to other criminals.

Rohit Chauhan, executive vice president of AI at Mastercard, says generative AI is adding some magic to fraud detection

When they have gathered thousands of card details, criminals will bundle them together and obscure all but four digits of the 16-digit card numbers. 

They will then advertise them on the dark web for other criminals to buy. 

This criminal activity falls under what UK Finance classes as unauthorised fraud – where the the card holder does not provide authorisation for a purchase and a transaction is carried out by a criminal.

Unauthorised card fraud losses totalled £551.3million in 2023 according to industry body UK Finance and amounted to 2.7million cases.

But now Mastercard is attempting to outsmart criminals by predicting which cards are compromised online so they can be blocked before criminals get to use them. 

Here’s how they are doing it.  

Mastercard scans millions of card numbers across its network and takes all of the ones it identifies as having potentially suspicious activity on them as well as all of the four digit adverts for stolen cards that they spot on the dark web.

Matching up these numbers up and filling in the gaps between the four digit card ads and the full 16 digit card numbers is where the generative AI comes in.   

Using its predictive powers, Mastercard is able to quickly find the full card details of the compromised cards.

Chauhan said: ‘It’s a little bit like trying to complete a jigsaw when you only have four pieces and don’t know what the pattern for the puzzle looks like. Generative AI helps to complete the puzzle by filling the missing numbers in.’

He added: ‘With these compromised cards it’s all about detecting them quickly and with high accuracy. This way we can get a head start on helping banks understand which cards are compromised so they can take action.’

The new technology scans transaction data across cards and merchants Mastercard partners with at faster rates than previously imaginable.   

Mastercard said the new tech is able to double the speed at which Mastercard can detect compromised cards and slash the number of false positives during the detection of fraudulent transactions against potentially compromised cards by up to 200 per cent. 

Generative AI will also help Mastercard speed up the identification of merchants which are at risk from or already compromised by fraudsters by 300 per cent. 

As a result, Mastercard said it will be able to alert banks more quickly and with greater accuracy when a card is likely to be compromised.  

The card can then be blocked for the customer and reissued. 

It has not been rolled out across all banks yet. With the new generative AI tool, Mastercard said it is providing a solution to banks in the UK but it is up to the banks whether they implement it.  

Mastercard said that banks could use the technology to monitor attempted transactions on the compromised card to mitigate fraud and enhance cyber security. 

Johan Gerber, executive vice president of security and cyber innovation at Mastercard said: ‘Until now fraudsters may have thought they were operating in obscurity, seeking to launder the card details of millions of unsuspecting victims. 

‘Thanks to our world-leading cyber technology we can now piece together the jigsaw – enhancing trust to banks, their customers and the digital ecosystem as a whole.’  

Chauhan added: ‘It’s not a silver bullet, no one thing is going to be able to stop fraud in its tracks. But this is great example of how we are innovating to stay ahead of the fraudsters.’

Visa launched Visa Protect which uses AI to scan billions of historic transactions and identify compromised cards

How Visa Protect is fighting fraud

Mastercard rival Visa has also recently announced it is using new AI technology to combat fraud in a partnership with Pay.UK

It has been using AI to analyse billions of historic bank transactions for signs of fraudulent activity. It successfully identified 54 per cent of the fraudulent transactions in addition to those already identified by banks. 

Called Visa Protect, Visa says this could save the UK over £330million a year on fraud and Authorised Push Payment (APP) scams.

In a pilot with Pay UK, Visa analysed billions of card transactions over the past 12 months using AI. 

Visa is now making its AI technology available to all banks in the UK. 

Visa says the new technology will help to intercept suspected fraudulent transactions in real-time, stopping scams before the money ever leaves the victim’s bank account.

It has the potential to claw back £330million from the hands of criminals of £600million lost annually to account-to-account scams and fraud over recent years in the UK according to UK Finance. 

Mandy Lamb, managing director of Visa UK & Ireland, said: ‘The UK has one of the most developed payment systems in the world, but also sees some of the highest levels of account-to-account fraud.

‘We are now bringing our AI capabilities to fight fraud and scams on account-to-account payments before they happen.’