NatWest Group fights fraud and improves customer experiences
NatWest Group (formerly RBS Group) uses Nuance voice biometrics technology to quickly detect fraud attempts coming into its call centre and disrupt organised crime activities across all its customer engagement channels.
Finding a secure, seamless experience
NatWest needed to put more effective security mechanisms in place while still delivering a fast, smooth experience for genuine customers. That meant finding ways to rely less on passwords and other static identifiers that can be stolen or forgotten. Find out how Nuance helped.

Increasingly sophisticated fraud attempts on the voice channel and beyond.
Nuance Gatekeeper analyses incoming calls to the contact centre, authenticating legitimate customers and identifying fraudsters in real‑time.
17 million calls screened, and 23 thousand fraud alerts created, leading to over 300% ROI in one year.
Protecting 19 million customers from fraud
Like any bank, the call centre is an important customer service channel for the NatWest Group, which serves 19 million customers across 12 banking and financial services brands.
But the voice channel is also a prime target for fraudulent activity. To combat criminal behaviour—both from lone‑wolf attackers and organised crime networks—NatWest was looking for ways to get a clearer view of fraud indicators across all its customer engagement channels.
The bank needed to put more effective security mechanisms in place while still delivering a fast, smooth experience for genuine customers. That meant finding ways to rely less on passwords and other static identifiers that can be stolen or forgotten, and it knew that voice biometrics could be an important piece of the anti‑fraud puzzle.
NatWest had been using Nuance voice biometrics in its Coutts private banking brand, so it had seen how effective the technology could be at providing a seamless experience for customers and protecting them from fraud.
Fighting crime in the call centre and beyond
NatWest deployed Nuance Gatekeeper on‑premises to screen every incoming call and compare voice characteristics (including pitch, cadence, and accent) to a library of fraudster voiceprints. The system identifies known fraudsters in real‑time, alerting the call centre agent and the fraud team.
NatWest has also enrolled legitimate customers to rapidly authenticate them based on the sound of their voice during their first few seconds of natural conversation with an agent. This creates a better experience for customers, who no longer have to remember passwords or security question answers; and it creates a better experience for agents, too, who can focus on actually helping the customers.
The Nuance solution also enables the bank to take a holistic approach to fraud detection and prevention. By combining data with information from other criminal activity detection tools, NatWest has discovered that fraudsters on the voice channel also perpetrate a lot of fraud on digital channels. Armed with that knowledge, the bank has been able to identify and disrupt organised crime activities to protect its customers and assist law enforcement.
calls screened
fraud alerts
ROI in one year
Saving money and disrupting organised crime
In less than a year, NatWest has screened 17 million inbound calls. Of these, 23,000 have led to alerts, and the bank has found that one in every 3,500 calls is a fraud attempt. Stopping fraudsters in their tracks is already paying off financially, as Jason Costain, the bank’s head of fraud strategy and relationship management, explains: “Although this initiative isn’t just aimed at reducing losses, we expected to save a reasonable amount of money, and we’ve already saved one and a half times that.”
“It’s not just about stopping financial loss—it’s about disrupting criminals,” says Jason. “For example, one prolific fraudster identified through Nuance was connected to suspect logins on 1,500 bank accounts. That’s helped us protect potential fraud victims and identify the ‘mules’ being used by the crime network to perpetrate fraud, leading to two arrests so far.”