Decisioning AI Targets Friendly Fraud

Such immense effort is expended fighting determined fraudsters that it’s easy to minimize friendly fraud — a frustrating state of affairs that leaves financial institutions (FIs) and merchants on the hook for purchases disowned, justly or otherwise, by legitimate cardholders.

PYMNTS’ September FI Fraud Decisioning Playbook sponsored by Simility, a PayPal service, explores pressing matters around chargebacks as eCommerce becomes the dominant retail channel at a time when every transaction is worth more than the sum of items in a cart.

“FIs handling chargeback claims must do their best to protect consumers and merchant customers, but this can be a difficult undertaking,” per the new Playbook.

“Resolving disputes can be time intensive, forcing banks to take on the administrative tasks required to gather and assess evidence. Card issuers that believe chargeback claims are valid then ask merchant acquirers to send funds on behalf of merchants to cover the transaction reversals — unless the sellers wish to dispute the claims. These processes can be expensive for FIs as well as retailers,” the Playbook states.

Meanwhile, “Acquirers facilitating chargebacks risk failing to receive compensation from merchants that become insolvent, and both issuers and retailers can become tangled in administrative work should the latter dispute the transactions.”

This is where artificial intelligence (AI) technologies like Adaptive Decisioning are proving invaluable. The new Playbook contains numerous use cases showing how chargebacks and frictions can be well-managed.

Spotting The Fakers

They call it “friendly” fraud because it isn’t always malicious, and divining the good actors from the bad can be difficult: just ask any call center rep who’s ever handled a chargeback call.

Data insights are the primary way that merchants and FIs are more successfully managing chargeback situations now, coupled with powerful decisioning platforms.

“FIs that firmly understand customers’ card-purchasing histories and patterns can better assess whether contested transactions match their typical behaviors,” per the Playbook.

“Issuers that spot unusual details or activities suggesting foul play could determine that chargebacks are justified, but transactions that appear normal could warrant reaching out to cardholders to jog their memories and ultimately result in erroneous claims being dropped.”

Superhuman data powers of fraud detection platforms are mitigating this more now, and “AI-powered tools can analyze consumers’ transaction histories as well as industry-specific trends and then determine whether contested purchases are likely legitimate or fraudulent.”

Informed Evaluations, Quickly

In a time of mass remote digital onboarding and nerve-wracking levels of account holder anonymity, too many fraudsters are easily penetrating outdated defensive systems.

There’s an urgent push to modernize across industries now, perhaps nowhere more so than among the various FIs and merchants on the front lines of society’s COVID-19 digital shift.

“FIs are … taking aim at the issue, leveraging advanced analytics solutions to help them better distinguish between genuine and illegitimate claims. AI tools can reportedly help banks review claims and draw on customers’ data to inform their evaluations,” according to the September FI Fraud Decisioning Playbook.

“This could allow FIs to determine whether shoppers previously patronized the merchants in question, for example, and whether disputed transactions fit preexisting behavioral profiles.”