Mastercard Says New AI Model Ups Fraud Detection by 20%

Mastercard has introduced a generative artificial intelligence (AI) solution to help prevent fraudulent transactions.

The company has upgraded its Decision Intelligence solution to let it “scan an unprecedented one trillion data points to predict whether a transaction is likely to be genuine or not,” per a Thursday (Feb. 1) press release.

The tool, dubbed Decision Intelligence Pro, examines the relationships between the entities involved in a transaction to assess its risk, Mastercard said in the release. It arrives as banks are increasingly using AI to ferret out criminal activity.

“In less than 50 milliseconds, this technology improves the overall DI score, sharpening the data provided to banks,” the release said. “Initial modeling shows AI enhancements boost fraud detection rates on average by 20 percent and as high as 300 percent in some instances.”

According to the release, the new version of AI will give banks greater ability to protect customers from fraud and help reduce false positives, or legitimate transactions incorrectly marked as fraudulent.

“The precision of the solution — achieved by scanning potential points of sale in real time — has been shown in our own analysis to not only increase accuracy, but also reduce the number of false positives by more than 85 percent,” said Ajay Bhalla, president of Cyber and Intelligence at Mastercard.

The announcement came one day after Mastercard released quarterly earnings that showed a 17% increase in its value-added services and solutions revenue, driven chiefly by demand for cyber and intelligence solutions, along with identity and authentication products.

Meanwhile, PYMNTS spoke with Michael Shearer, chief solutions officer at HAWK:AI, in an interview posted Thursday about the use of artificial intelligence to fight the multi-trillion dollar problem of fraud and other financial crimes.

“It is essentially an adversarial game; criminals are out to make money, and the financial community needs to curtail that activity. What’s different now is that both sides are armed with some really impressive technology,” Shearer said.

“It is incumbent on everybody in the network to use this technology as effectively as they can and play their part in driving up the difficulty and the risk of abusing the financial system such that it becomes prohibitive and the reward is not worth it for criminals,” he added.

Research by PYMNTS Intelligence has shown that employing new technology can pay off for banks, as 66% of financial institutions that used AI or machine learning (ML) saw an overall decrease in fraud rates.