LendingClub attributes fraud detection capabilities to its proprietary AI-driven system.  

“Our fraud rate is near zero, less than 3 basis points,” Lukasz Strozek, chief technology officer of LendingClub, told FinAi News. “It’s the product of years of investment in AI-powered fraud detection and a layered defense system that looks at hundreds of real-time signals for every transaction.” 

Implemented “several years ago” the fraud defense system is built and managed by LendingClub’s engineering and credit risk teams, Strozek said. 

The fraud defense system combines in-house machine learning models with more than 20 external data integrations, Strozek said. The system can analyze more than 400 data points from device intelligence, identity networks and behavioral analytics. 

“We designed it to be fast, predictive and continuously learning,” Strozek said.  

Although the system pulls from external data sources, the architecture and decisioning logic are developed internally, he said, noting that decisions can be made in less than a second. 

“The fraud system has been in place for several years, refined continuously with the help of our growing dataset and feedback loops from millions of lending interactions,” he said.  

Lending fraud rates

LendingClub says it is keeping fraud rates low at a time when fraud losses across mortgage and consumer lending are mounting. 

The National Mortgage Application Fraud Risk Index for the third quarter found that fraud risk is up 8.2% year over year, with an estimated 1 in 118 applications showing indications of fraud.  

In 2024, consumers reported losing more than $12.5 billion to fraud, a 25% increase YoY, according to FTC data. 

With AI in place, LendingClub is keeping fraud at bay, Strozek said. “It’s one of the biggest reasons we’ve maintained such consistently low fraud losses across all economic cycles.” 

Register here for early-bird pricing for the inaugural FinAi Banking Summit 2026, taking place March 2-3 in Denver. View the full event agenda here.