Plaid Study Links Account Checks to Faster Fraud Detection

A new August study from PYMNTS Intelligence and Plaid reveals that 47% of surveyed companies struggle to detect suspected fraud either immediately or within a short time frame, despite employing advanced defenses. While 90% of firms utilize rules-based scoring, 83% employ machine learning or predictive risk scoring, and 81% examine payment histories, the issue lies in the outdated nature of these controls designed for slower payments processing.

Real-time payments have shifted the dynamics, making traditional fraud detection methods obsolete as settlement occurs instantly and irrevocably. A majority (57%) of firms report an increase in fraud attempts compared to the previous year. In terms of payment flows, incoming customer payments are perceived as the most risky by 47% of surveyed businesses.

The urgency in fraud prevention has shifted from simply identifying suspicious activity to detecting it before the money becomes irrecoverable, highlighting the strategic importance of real-time detection. Despite the operational benefits, real-time payment adoption remains cautious, with only 24% of firms having fully adapted their fraud systems for real-time payments, even though 67% note an increase in fraud exposure due to this transaction type.

Firms that verify account ownership in real time detect suspected fraud 60% of the time, compared to just 39% for those who do not verify in real time. Moreover, real-time verifiers are more than twice as likely to identify fraud before funds are transferred (21% vs. 10%). While account verification alone does not guarantee faster fraud detection, it suggests an architectural advantage for firms that can link identity, ownership, and account information directly to transaction decisions.

The study indicates that many businesses already possess the necessary infrastructure for real-time fraud detection but are not fully leveraging it.