Collaborative Defense: The Role of ‘Intelligent Friction’ and AI in Fraud Prevention

Every now and then, an executive in the payments business comes up with a catchphrase that captures a complex concept, and you know it’s going to resonate.

A good example lies in describing the tension between tight security protocols and practices and a positive customer experience. The catchphrase that nails it is “intelligent friction.”

Attribute that one to Graeme Bullock, EMEA sales leader at financial verification company Entersekt. He turned the phrase during a conversation with PYMNTS and a panel that included Nordic payment processor Nexi Group and J.P. Morgan, focusing on the tension mentioned earlier.

Bullock introduced the concept of “intelligent friction,” which emphasizes applying security measures based on the level of risk associated with a transaction rather than a one-size-fits-all approach. This approach ensures that security measures are dynamic and context-specific, reducing unnecessary interruptions for legitimate transactions while effectively targeting potentially fraudulent activities.

As Bullock explained it, intelligent friction involves analyzing multiple factors, including the device used, location, transaction type and user behavior, to assess the risk level accurately.

“Rather than looking at a single point and making a decision based on a rule, it’s more about understanding the full context of the interaction,” he said. “We need to consider the omnichannel approach. Sometimes, I may interact with my financial institution on my laptop, and other times through my mobile app. The institution must make decisions based on my identity regardless of the channel. This requires a seamless integration of data from various sources to create a holistic view of the customer’s behavior.”

The panelists discussing the issue agreed that intelligent friction is crucial in maintaining a balance between security and user experience. As Nexi Group Head of Risk Management Services Sean Neary added, the consumer knows what they are doing when authorizing a transaction. Payment systems, therefore, need to be smart enough to introduce the right level of intervention without disrupting legitimate activities.

However, the picture gets a bit more complicated when larger commercial transactions are in play. As J.P. Morgan Head of Fraud Prevention for Commercial Banking Alec Grant told the panel, in commercial payments, clients are sometimes a few steps removed from personal knowledge of who they’re paying. He applied Bullock’s phrasing to his experience.

“Our friction involves having conversations with clients to ensure they understand the risks and verify the transaction’s legitimacy,” Grant said. “We train our teams in psychological profiling to listen and challenge appropriately. This approach has significantly reduced fraudulent transactions without compromising the client experience.”

The Consortium Approach

Intelligent friction, therefore, represents a nuanced and sophisticated approach to fraud prevention, ensuring that security measures are as seamless as possible while effectively mitigating risks. By using advanced technologies and fostering cross-industry collaboration, financial institutions can protect their customers and maintain trust.

That cross-industry collaboration was an important theme during the panel discussion.

It’s imperative that we standardize data sharing and classification,” he said. “This ensures that the consortium model is effective in preventing fraud across institutions.”

Neary also highlighted the role of privacy-enhancing technologies in enabling secure data sharing without compromising personal information.

“These technologies allow us to tokenize and standardize [personally identifiable information (PII)] data, facilitating secure and meaningful data sharing across the consortium,” he explained.

The consortium model has worked in specific areas of fraud. For example, Bullock underscored the success of collaboration in reducing authorized push payment (APP) fraud. He said the contingent reimbursement model introduced in the United Kingdom has plateaued the increase in APP fraud, proving to him that collaboration and data sharing are essential in tackling fraud.

APP fraud has emerged as a concern, particularly in Europe and the U.K. Grant highlighted the complexities of this fraud type, where customers are tricked into authorizing payments to fraudsters.

“We are seeing a two-thirds reduction in the clients letting the funds go just by setting up these very specific teams, who are trained to handle these situations,” he said. “It’s fantastic for our clients because they appreciate the extra layer of protection.”

The AI Angle

Artificial intelligence is also playing an increasingly vital role in fraud prevention. Bullock emphasized the importance of AI in creating a multilayered approach to security.

“AI helps us analyze the behavior of an individual and a fraudster,” he said. “We use risk modeling for that, considering factors like location, IP address and device type. This allows us to make informed decisions about the legitimacy of a transaction.”

Grant echoed this sentiment, highlighting the precision AI brings to fraud detection.

“In the last two to three years, we’ve worked closely together to apply AI in identifying fraud,” he said. “We are seeing a significant reduction in interrupted transactions while increasing fraud detection.”

Looking into the future, expect to hear a lot more about intelligent friction and the consortium approach. As financial institutions strive to stay ahead of fraudsters, adaptability and collaboration will remain key. Neary emphasized the importance of a layered, configurable platform that can adapt to different stages of digitalization globally.

“Choose your battles, connection points, and invest in technologies that offer harmony in layered security, ensuring customer satisfaction and trust,” he advised.

Grant shared his vision for the future, where data sharing among banks could reduce fraud losses.

“If we could share information anonymously with other banks, we could collectively make a massive difference in reducing fraud losses to clients,” he said.

Bullock said there is a need for continuous evolution in fraud prevention.

“We’re never going to get to 100%, but by adopting a multilayered, context-aware approach, we can make the best decisions to protect our customers while ensuring a seamless experience,” he concluded.

PYMNTS-MonitorEdge-May-2024