Credit One Executive Talks Fraud Prevention Technology at SAS Innovate Conference

Long Jiang, Assistant Vice President of Fraud Analytics and Strategy at Credit One Bank spoke at the SAS Innovate conference this month. He and SAS’ Stu Bradley, Sr. VP of Risk, Fraud, & Compliance Solutions, discussed the common challenges anti-fraud professionals face in the generative AI age and how breakthroughs in technology have helped improve Credit One’s services.

Among the paramount industry challenges, Jiang cited advancements in AI-driven social engineering. Fraudsters are using deepfake technology, including customer identity deepfakes, to manipulate voice and facial recognition systems. Generative AI tools are also helping bad actors polish their phishing ploys. AI-powered content generation is helping reduce grammatical errors and misspellings in scam texts and emails, enhancing their believability.

GenAI is also fueling rampant growth in synthetic identity fraud, creating fake identities and blending real and fabricated data to bypass verification processes. It driving advanced scams such as "man in the middle" attacks. On the government front, AI tools are being exploited to file fraudulent tax returns using stolen or synthesized personal information.

Jiang shared a well-known quote from Warren Buffett: “Only when the tide goes out do you discover who’s been swimming naked.”

“This underscores the importance of prudence in financial strategies to avoid high fraud losses,” said Jiang.

Fighting fraud, AI vs. AI

While fraudsters are becoming much more sophisticated, banks can use the speed of technology to keep pace in fighting fraud, Jiang said. Real-time capabilities delivered on an integrated, cloud-based platform helps facilitate real-time decisions and enhance modeling, which in turn reduce false positives and increase fraud detection rates. Technology is helping Credit One drive higher revenue, improve customer service and lower declines.

Jiang said that another big positive for banks — and specifically credit card companies — is that AI and machine learning is helping with customer experience and security by speeding up the process of disputes.

Consumers universally value convenience, security, and tangible benefits, he said. These needs align closely with what fraudsters target, making robust security measures critical in areas such as:

• First-Party Fraud Detection: Identifying fraudulent activities initiated by the account holder.
• Synthetic Identity Detection: Spotting fraud applications made using fabricated identities.
• Payment Validation and Recovery Opportunities: Ensuring payments are legitimate and recovering funds in fraudulent cases.
• Authorization Detection Systems: Monitoring and verifying transactions to prevent unauthorized access.

The bigger picture: a differentiated customer experience

Fraud-fighting goes hand-in-hand with other operational improvements at Credit One, said Jiang. The bank is putting emphasis on creating synergy across departments to reduce net losses and other risks — but the bigger picture is to wow the bank’s customers.

SAS has helped the Credit One develop capabilities, from application to transaction. Creating tailored customer profiles helps Credit One differentiate normal customer behavior from potential fraud activities. Continuous monitoring helps the bank track activities like authorization velocity, account inquiries, and peer-to-peer transactions to identify discrepancies.

Once fraud is detected, the operations team collaborates with merchants to recover fraudulent funds and expedite the claim process for customers, recognizing the frustration involved in such situations and aiming to minimize inconvenience.

“Customer service is dramatically better than 18 months ago,” Jiang said of Credit One’s SAS investments.

Impacts across the enterprise

Financial firms will need technology to navigate the challenges ahead. Jiang noted that US commercial banks saw a rise in credit card delinquency rates to 3.1% quarter over quarter.

To make the most significant impact with technology, Jiang advised that data analysis needs to be implemented across departments and communicate as one. Also, the more data, the better the AI.

Bradley agreed with Jiang. “Some financial institutions are still looking at risk in silos. They need to look holistically, which will result in better risk modeling.” He mentioned that 80% of fraud investigators’ time is spent collecting data information that Gen AI can do for them, which would result in more efficiency and a better job experience.

Jiang and Bradley also discussed how stronger data analysis goes beyond just the benefits of fraud prevention and customer service. Clearly the banking crisis in early 2023 exposed the need for a deeper look at data analysis as the first inflationary environment in decades exposed inefficient risk modeling.

SAS Fraud Management uses data analytics and machine learning to monitor payments and nonmonetary transactions to help detect, prevent and manage fraud for financial institutions.