Payments and financial services providers are facing increasing pressure to strengthen fraud defenses while maintaining seamless customer experiences. To pull off this difficult but critical balancing act, providers must take a multifaceted approach to risk mitigation, leveraging both highly strategic practices and powerful anti-fraud technology. 

In a recent webinar hosted by Payments Dive, Galileo Financial Technologies Head of Global Payments Risk Management Max Spivakovsky discussed how providers can successfully navigate the fast-evolving fraud risk landscape while preserving the positive user experiences today’s consumers expect. 

Customers Want Security and Convenience

The total value of fraudulent card transactions reached a staggering $33.8 billion in 2023, according to the Nilson Report. Meanwhile, card not present (CNP) fraud is surging, increasing by 35 percent in 2023 and costing businesses an average of $3.75 per dollar lost, a LexisNexis study found.

Beyond direct losses, fraud significantly impacts operations, with businesses spending 23 percent of their operational costs on fraud prevention and recovery according to the Ravelin Fraud Report.

3 Steps for Optimizing Financial Fraud Detection Tools

As those statistics make clear, the stakes have never been higher for payment providers in the ongoing fight against fraud. 

However, the numbers also reveal that while robust fraud prevention is essential, it cannot come at the expense of a smooth, seamless customer experience; 67 percent of consumers are even willing to abandon digital payment transactions due to overly complex authentication, Experian found. 

“The customer friction is always a big pain point,” noted Spivakovsky. “Within fraud [defense], we are always trying to balance between the experience while at the same time mitigating and detecting fraud.”

Risk-Based Authentication 

Spivakovsky highlighted Risk-Based Authentication (RBA) as a leading strategic approach for balancing security with customer experience. RBA adjusts authentication levels based on real-time risk signals gathered from multiple payment channels and vendors. By evaluating patterns such as locations, devices and network details, RBA can tailor friction levels for individual users in real-time, deploying stringent authentication requirements only when necessary.

“It's a dynamic approach,” Spivakovsky said of RBA. “It basically allows us dynamically to adjust certain levels of authentication based on the risk signals as we are getting them from our multiple channels, multiple channels of payment and multiple vendors.”

The Executive’s Guide to Comprehensive Payments Fraud Detection

That means that, when an attempted transaction breaks an established pattern or displays some kind of anomaly, additional authentication procedures can be automatically introduced–as can further investigation into the consumer or the transaction itself to ensure that it's legitimate, added Spivakovsky. 

Meanwhile, transactions that aren’t flagged proceed as normal, preserving a fast, seamless experience for the vast majority of legitimate customers. 

AI and Machine Learning

Along with an RBA-based approach to mitigation, Spivakovsky also cited AI and machine learning as two particularly effective tools to fight fraud without hindering user experience. 

AI/ML can process enormous transaction volumes to identify risks with a high degree of precision. That precision is vital to detecting fraud amid a massive and ever-growing number of transactions; the Automated Clearing House network alone processed more than 1.2 billion transactions in 2024–and reducing false positives.

“The latest evolution of AI and ML allows us to be much more proactive in our real-time analysis and real time pattern recognition,” said Spivakovsky. “In real-time, we are able to make decisions that [mitigate] financial implications of fraud.”

How Banks Are Leveraging Anti-Fraud Tech to Fight Fraud

Machine learning's adaptive capabilities are particularly valuable, the Galileo risk head noted. “The ability of ML to continuously learn from the data, keep improving itself and adapt to the new fraud factors, while at the same time taking the facts of the real-time analysis, is what makes… this a very robust methodology,” he said. 

Spivakovsky added that the latest evolution of AI allows for more proactive real-time analysis and pattern recognition, while machine learning continuously improves accuracy by adapting to new fraud factors. This adaptive learning capability helps systems identify previously unseen attack vectors, he noted..

Want to Learn More? 

Watch the full webinar for more on payment security measures that protect both the bottom line and the customer experience.