Contact centers give banks a unique opportunity to interact with their customers. Get it right, and they can win a customer for life. But banks have to perform a delicate balancing act. They want to enable frictionless and convenient customer experiences while simultaneously protecting against a broad range of threats across a variety of channels and locations. Safeguards are necessary to keep criminals at bay, but too many barriers risk poor experiences that alienate customers at a time when it has never been easier to switch banks.
Fortunately, the right security tools can help. Authentication and fraud prevention are two sides of the same coin. Artificial intelligence (AI) can give customers frictionless authentication that immediately recognizes them and greets them by name. The same AI can prevent fraudsters seeking to drain a customer’s account.
Traditional call fraud monitoring
In smaller firms, customer service agents are often the front line of defense. In lieu of dedicated resources to monitor whether a caller is fraudulent, agents are given responsibility for monitoring. This asks a lot of agents, as fraudsters are becoming more and more sophisticated.
At larger banks, fraud monitoring is conducted by dedicated fraud teams. However, this doesn’t make the task any less challenging. Pindrop research suggests that as many as 2% of total calls can be fraudulent. With tens of thousands of calls a week, identifying those fraudulent calls is like finding a needle in a haystack.
AI helps solve these problems. Using machine learning and deep neural network technology, AI models can be trained on vast datasets of phone network, device and even voice characteristics. AI can learn how to spot the high-risk activities that indicate potential fraud. Over time, the models learn from real-world feedback, eventually allowing them to predict fraud with a high degree of accuracy.
For smaller banks, AI frees customer service agents from the burden of real-time voice fraud monitoring. As a result, agents can get on with what they are actually trained to do: helping customers solve problems. With AI-based voice fraud monitoring, smaller banks are able to increase security and improve customer experience (CX) by eliminating the need for agents to constantly engage in fraud detection.
More sophisticated organizations can also benefit. Omni channel monitoring allows businesses to use insights from one place, like the contact center, to spot that an account is at risk and prevent fraud attempts somewhere else, like on the website.
Driving CX through security
Historically, the balance of offering frictionless customer service while also mitigating risk has had to err on the side of caution and cumbersome security measures. Fraudster toolkits have grown in sophistication. Criminals are able to access growing volumes of stolen consumer data on the black market. When combined with technologies like caller ID spoofing, voice synthesis, deepfake and robotic dialling technologies, fraudsters can leverage this stolen data with ruthless efficacy.
However, AI done right can turn security and CX into a virtuous circle. The better banks are at spotting fraud and authenticating customers, the easier it becomes to support self-service capabilities, such as Interactive Voice Response (IVR), apps and online portals, which in turn drive better customer experiences.
Self-service tools, such as Interactive Voice Response (IVRs), can provide a great customer experience, but can also be ripe for misuse by criminals. It is easy to exploit an insecure IVR to conduct deep reconnaissance about a targeted account. By the time the fraudster reaches a contact center agent, they will have gained enough intelligence to present themselves as a genuine customer and take over that customer’s accounts. Given what’s at stake, any bank offering IVR services would traditionally need to dedicate significant security resources to shore up the service.
AI makes securing IVRs easy. Models can be configured to monitor self-service channels in real time so that surveillance and reconnaissance attempts can be identified and flagged to fraud teams before they’re lost. In fact, Pindrop's latest research indicates that by monitoring suspicious activity in the IVR and matching these alerts to the related accounts, organizations can predict fraud several weeks in advance and help ensure that vulnerable accounts are protected.
AI for the future
The cases for AI in security are many and varied. The above examples reveal a common truth: with AI, banks can offer better customer service while significantly enhancing the efficacy of their security. AI will help enable the next generation of customer-centric services. It’s time to get on board.
This article originally ran in Security, a twice-monthly security-focused eNewsletter for security end users, brought to you by Security Magazine. Subscribe here.