How Visa’s AI Prevented Billions of Dollars in Fraud

Visa, the largest payment processor globally, has dedicated approximately $500 million to developing artificial intelligence (AI) and data infrastructure, improving its ability to prevent fraudulent activity. Charles Lobo, Visa Regional Risk Officer for Central and Eastern Europe, Middle East, and Africa, reported that these advancements have helped Visa intervene against "80 million potential fraudulent transactions," which may have otherwise resulted in about $40 billion in fraud in 2023.

Visa's AI systems can help prevent fraud by analyzing every transaction and identifying patterns and anomalies. It can raise red flags for transactions that deviate from a customer's typical behavior, like using their card abroad or making purchases that are out of pattern, prompting immediate alerts to both customers and card issuers.

Speaking with CNBC last month, James Mirfin, Global Head of Risk and Identity Solutions at Visa, shared that the AI evaluates the likelihood of fraud by considering "over 500 different elements," incorporating factors such as the purchasing device, location specifics, merchant identifiers, and transaction amounts.

“We score that and we create a score –that’s an AI model that will actually do that. We do about 300 billion transactions a year,” Mirfin said.

Mirfin said that enumeration attacks, where criminals use automated tools to generate and test multiple card numbers until they find a valid one, can cause $1.1 billion annually in losses. Visa's AI can detect and block them before they cause any damage.?He also told CNBC that the amount of fraud stopped by Visa's AI systems had doubled from 2022 to the same period in 2023.?