Locking out fraud without increasing false positives can seem nearly impossible for financial institutions (FIs), but it doesn’t have to be. In the Preventing Financial Crimes Playbook, Synchrony Chief Information Officer Bess Healy tells PYMNTS about the company’s path to 90% accuracy with the help of artificial intelligence (AI) and machine learning (ML).
The pandemic accelerated the rise of digital-first consumers — those who do most or even all their banking and shopping online.
This has created opportunities for financial institutions (FIs) to expand their digital businesses, but also has intensified the risks. Online fraud and other financial crimes are soaring, and FIs need to detect cybercriminals and block their attempts without risking false positives and intrusive security checks that derail legitimate customers’ transactions. Artificial intelligence (AI) systems can help FIs overcome these obstacles, according to Bess Healy, senior vice president and chief information officer of Synchrony, who spoke with PYMNTS in an interview.
Synchrony’s long-term investments in technology have included AI and cloud systems as well as customer education about digital services. These bore fruit during the pandemic, Healy said, positioning the company to meet its customers where they were in the rapid digital shift. AI systems also enabled Synchrony to enhance and personalize customer experiences despite unprecedented disruptions to both consumers’ lifestyles and the wider business environment.
“[AI should be viewed] as an entire ecosystem of automation,” she explained. “[Its applications range] from process and workflow automation to really changing what our employees are doing so that they’re not [performing] repetitive [tasks]. … We’re applying intelligence to the process … to have our employees do things that we want only humans to do.”
Synchrony’s AI umbrella covers three distinct components, Healy explained: intelligent automation, an intelligent virtual assistant and machine learning (ML). Intelligent automation includes AI-powered automation capabilities that support both front- and back-end aspects of the business. Intelligent virtual assistant functions mainly interact with customers and support internal operations. ML technologies drive the overall learning and evolution of the AI system.
AI-Powered Fraud Prevention Helps Ensure Frictionless Customer Experiences
Synchrony’s AI ecosystem is proving extremely adept at accurate fraud detection. The system employs ML and automated decision-making to learn from and adapt to new behavioral patterns. This is critical because of cybercrime’s swiftly changing nature.
“The bad guys never sleep,” Healy said. “The pace at which change is coming at us makes it crucial to employ ML and leverage and harness the data we have to feed those capabilities.”
Synchrony currently is achieving greater than 90% accuracy in detecting transactions’ legitimacy, and Healy said she is optimistic that this can be improved even further. She emphasized that effective fraud prevention includes avoiding false positives, which is fundamental to both monitoring efficiency and making the customer journey as frictionless as possible. False positives and other security warnings are stressful for consumers and often cause them to give up on making their desired purchases. Blocked transactions also cost businesses lost sales and leave bad impressions with would-be customers.
AI-Based Credit Decisioning Boosts Sales and Reduces Losses
AI also plays a key role in Synchrony’s lending operations through Prism, the company’s credit decisioning system.
“[AI] has truly revolutionized the way we do business on the credit decisioning front,” Healy said, “[by] leveraging [our] expanded data footprint to give us a much more holistic perspective of the customer and [allowing us to] make smarter decisions.”
This is central to the know your customer (KYC) process, which also helps reduce fraud and leads to higher approval rates, improving sales for merchants and avoiding the failed transactions that ruin the customer experience. Accurate credit decisioning also reduces the long-term credit risk of customer defaults, and fast and accurate approvals of credit applications and transactions keep the process smooth from the customer’s side, too.
Synchrony is employing AI in other areas as well, Healy said, including speech analysis to gather insights about customer sentiment based on data, such as speaking tone and pace. This helped the company gauge how its pandemic relief programs were affecting customers, for example.
“We use speech analytics to understand … a potential concern around compliance or a complaint,” she said. “We want, over time, to continue to expand our use of data, including speech analytics data, to be more personalized in our interactions with customers.”
The increasing sophistication of AI makes it very important to use the technology ethically — an approach Healy calls “responsible AI.”
“[This] means that we’re focused on doing the right thing for our customers and our partners, that we’re compliant from a legal and regulatory standpoint [and] that we’re harnessing the capabilities and the data in a way that reflects our carrying value,” she explained.
Synchrony has already developed a robust and effective AI ecosystem, but the company has no plans to slow its search for new use cases for the technology.
“We are very happy with the results — but we’re not satisfied,” Healy said. “We’re always focused on use improvement, both for customer experiences and for better outcomes for our partners. So, we will continue our investment and focus on maturity in [existing] spaces, on new tool sets and capabilities and on further harnessing our data.”