From peer to peer payments to real-time international transfers, the finance sector is growing rapidly with a blistering urgency of meeting customer demands. Open banking has unlocked a whole new realm for businesses to exploit the bank data and create innovative financial services. However, payment fraud is still a growing problem, which organizations have somewhat failed to eradicate. As mentioned by The Guardian, “Scammers have stolen more than £500m from UK bank customers in the first half of 2018.”
The number of fraudulent transactions, data breaches, and instances of identity theft continue to rise as fraudsters and hackers have become more sophisticated. Another report by Javelin Strategy and Research states that “16.7 million US adults experienced identity fraud in 2017, marking an 8% increase from 2016.”
The use of Artificial Intelligence (AI) and machine learning, to detect and prevent fraudulent activities is gaining traction among banks and FIs. Many banks have started to deploy AI as a direct response to combat frauds. Here are some findings from a recent survey, AI Innovation Playbook, published by PTMNTS in collaboration with Brighterion:
- 72.7% of the firms are currently using AI for payment fraud detection.
- 63.6% of FIs believe that AI is an effective tool for stopping fraud before it happens.
- The majority (80%) of fraud specialists believe AI-based platforms are effective in reducing payment frauds.
Visa prevents approx. $25 billion in fraud using Artificial Intelligence
By using Artificial Intelligence, Visa Inc. helped financial institutions to prevent an estimated $25 billion in annual frauds. Visa achieved this with the help of Visa Advanced Authorization (VAA) a comprehensive risk management tool based on AI to monitor transactions on VisaNet in real-time. The tool processed more than 127 billion transactions in 2018 so FIs can approve legitimate transactions quickly while detecting and preventing frauds.
VAA employed AI and neural networks to analyze 100% of the transactions in about one millisecond. In that millisecond, the AI searches for indicators of fraud — looking for activities and patterns common in fraudulent transactions.
However, Visa Inc. is not alone in using AI and ML to fight fraud. Mastercard has also used AI technology on a global scale to fight frauds. The company introduced Decision Intelligence, a decision and fraud detection system to increase the accuracy of approvals of genuine transactions and reduce false declines.
KPMG is another big name on the list of organizations exploring technology in order to prevent fraudulent activities. Recently, KPMG and Nets, a digital payment service provider in Europe, have collaborated to develop a fraud monitoring and prevention system – Nets Fraud Ensemble. The system will be AI-powered that will deploy machine learning to detect fraud indicators in real-time. It will also consist of multiple models working together to evaluate each transaction in ten milliseconds. The solution is believed to decrease operating costs and chargebacks, and balance accuracy with customer convenience.
Why AI is ideal for fighting payment frauds
Here’s a fact – VisaNet processed more than 127 billion transactions in 2018 alone, and the number is constantly rising. With billions of transactions being processed every second, AI is ideally suited to provide businesses with the speed and scale to take on the challenge of payment frauds. AI makes it possible, verify, and authenticate a huge amount of data in seconds. Moreover, AI’s ability to analyze trend-based data from supervised and unsupervised machine learning algorithms have proven to reduce the incidence of payment frauds. Here’s why we believe AI could be perfect for fighting payment frauds:
- AI is ideal for finding anomalies in huge data sets within seconds. Implementing technology, organizations can gain a greater level of accuracy and predictability in various fields, for example, consumer behaviour.
- Payment frauds have become more complicated and sophisticated. They are growing in complexities with different digital footprints, making them nearly undetectable using rule-based models. AI, coupled with unsupervised machine learning, can help find previously unknown patterns in data set without pre-existing labels.
- AI offers organizations the speed and scale to combat payment frauds. It gives them an immediate advantage in fighting against frauds and preventing them from happening. BNY Mellon (Bank of New York) implemented an AI-based fraud detection system last year to recognize patterns in millions of transactions and identify those that are likely fraudulent.
Banks and payment processors are committed to providing faster, simpler, and more secure services to the customers. As a result, they are increasingly using AI and ML to craft strategies that perform against evolving fraud tactics. If properly applied, AI can help provide the foundation for highly substantial fraud detection systems to understand evolving transaction patterns in real-time.