Cyberattacks are on the rise, with attempted payment fraud in the buy now, pay later segment seeing a 211% year-over-year increase, attempted payment fraud in fintechs was up 13% in 2022, and payment fraud losses are anticipated to grow 17% YoY to $48 billion by the end of this year.  

The increases were revealed in the “Q1 2023 Digital Trust and Safety Index” report released last week by fraud prevention platform Sift, which surveyed 1,000 U.S.-based consumers and analyzed its network of more than 34,000 websites and apps.

The jump in human-powered fraud is due to lack of hands-on handling of know your customer (KYC) processes and lack of appropriate funding in security, Meng Liu, senior analyst at research firm Forrester, told Bank Automation News. 

Fraud ROI to blame 

The report also found: 

• Card-not-present fraud is expected to cost digital merchants $9.49 billion this year; 
• Marketplaces saw average fraudulent transaction values jump 52% to $5,149; and 
• 16% of consumers have committed or know of someone who has taken part in payment fraud. 

“Fintechs are also doing a lot of financial incentives for customer acquisition, like rewards and cash back for new customers,” Liu said. “Fraudsters can also leverage bots and synthetic identities to register massive new accounts to take advantage of financial incentive and promotion programs.”  

The rise in fraud is largely due to the massive return on investment for social engineering scams, such as a fraudster pretending to be a bank official, are increasingly prevalent, said Jane Lee, trust and safety architect at Sift. 

“What [criminals] do is coerce [the victim] into sending over their verification code to access [the victim’s] bank account to commit the fraud,” Lee said, adding that the challenging part of mitigating this is that the victim is authorizing entry into their bank account. “We’re in an adversarial space,” she said. 

Fraud-mitigation tools

However, machine learning (ML) can be deployed to mitigate fraud. For example, Sift’s ML-powered Digital Trust and Safety platform can take large-scale data and feed it into the platform, providing insights and identifying fraudulent patterns, Lee said. 

“We recently announced that we ingest 1 trillion events annually in the platform … and so that gives teams the tools to be able to make more educated decisions based on pattern behavior or fraudulent patterns that we know to be suspicious,” she said. 

Further, ML when combined with AI helps financial institutions detect and stop fraud in real time, Liu said. Other useful technologies include biometrics authentication, behavioral biometrics authentication and multifactor authentication that allow users to ensure multiple steps have been taken to prevent suspicious activity, Forrester’s Liu added. 

This can “avoid the weakness from static passwords,” which are easily attacked by fraudsters, he said. The increased secure data sharing with external organizations by using privacy preserving technologies allow a better view of some group fraudsters’ accounts and device information. 

Mobile protection, biometrics

Forrester predicts that card-based payments will decline this year due to tokenization and 3D Secure 2.0 protocols that require authentication by the user if the risk of a fraudulent transaction is high, Liu said. 

“Higher consumer adoption of mobile payments will also keep many fraudsters at bay, as mobile devices can support one of the smoothest and strongest customer authentication processes: biometrics,” he said.