The Future of Online Banking Security
Fraud is the number one crime in the UK. Last year criminals stole £1.2bn from individuals through banking fraud and scams. In 2025, fraud had risen to around 4.2 million incidents, up 31%, according to the latest government ONS data, highlighting that more needs to be done to protect users’ data and money. This upward trend highlights the growing sophistication of online fraud. Cybercriminals are no longer relying on generic phishing emails sent to a wide audience in the hope of catching a few vulnerable victims. Instead, AI is enabling a far more targeted and convincing approach and deepfakes allow cybercriminals to produce convincing video and audio impersonations, making scams hard to detect and far more damaging to organisations and individuals.
Both customers and financial institutions are left bearing the fallout from these attacks. The consequences are felt far beyond financial loss. For individuals these scams can be deeply distressing, the emotional toll of being deceived and the hours spent recovering accounts or proving their own identity can be overwhelming. For institutions, fraud is costly. It’s not just the cost of reimbursing stolen funds, but the cost of investigating fraud cases, and reputational and brand damage.
The current landscape
To counter this approach by criminals, banks across the UK, US and Australia have adopted single phrase voice biometrics as a fast and efficient way to authenticate customers accessing telephone customer services and online banking. But, with the advances in deepfakes and voice cloning, a simplistic one-shot identity verification solution is already all but obsolete. As OpenAI’s CEO, Sam Altman said, ‘AI has fully defeated voice authentication alone as a form of identity check.’
This process assumes that once you have passed this verification test you remain the same caller. That’s a big assumption. As such we have now seen the adoption of passive verification security methods, which listen in to the call in the background to overcome the limitations of single phrase voice biometrics. However, passive verification also misses the point that voice is a multi-factor credential containing both the biometric of the speaker and the intent (or message) of the individual.
To prevent fraudsters from accessing someone else’s bank account, security needs to be more than just voice or a one-time password (OTP). To truly verify the person is who they say they are, financial institutions need to be adopting a fused biometric verification.
The solution
Imagine if banks could both see and hear its customers, continuously verifying them in real time. At FARx, we’re revolutionising the delivery of secure online services. Our proprietary machine learning technology is designed to recognise humans the way humans recognise each other and ultimately protect against fraudsters by combining voice, speech and face recognition to verify who is really on the end of each sentence.
Unlike traditional authentication methods that only verify identity at login, FARx provides seamless, frictionless and continuous multi-factor biometric recognition of the user. Not only does our technology analyse both the unique vocal traits and facial features, ensuring that access is granted only when the person is genuinely who they say they are, it can also recognise when someone is looking over the user’s shoulder, blurring sensitive information or even shutting down the session.
FARx is AI that knows you. The more it learns about your customers’ voices, facial expressions, and language behaviour, the easier it is able to identify if it is NOT them (i.e., when it is a fraudster attempting to masquerade as your customer). It also has the capability to recognise a user’s emotional state such as whether they appear stressed, worried or nervous. These emotional signals can highlight when a user is under duress or being coerced, promoting protective actions to prevent fraud.
Not only does FARx verify customers swiftly, it also flags potential fraudsters for further investigation, creating a ‘global fraud list’. It can capture the fraudster’s voice and face biometrics and either create a new profile for the criminal or match it against existing biometric data. This approach not only protects against known threats but also tracks new suspects, ensuring a robust defence against emerging actors.
With fraudsters evolving and a surge in financial fraud, traditional security methods simply aren’t enough. Static passwords, one-time codes, and MFA checks are increasingly vulnerable to sophisticated attacks. It’s time to move beyond outdated authentication and embrace a future where security is seamless, continuous, and truly personal.
Here at FARx, we’re the future of human and computer interaction. To learn more about how fused voice-face biometrics can be used to prevent and track fraud get in touch with our expert team here.
—