Organisations are now at the forefront of a battle against AI-driven voice fraud, as FARx releases an upgraded version of its biometrics software. The software is designed to address the challenges posed by advanced text-to-speech and voice cloning technologies, which have become adept at mimicking human voices. Traditional voice biometrics, previously used for authenticating human speech, struggle to detect this newer AI-generated voice technology.
Significantly, recent figures highlight a drastic increase in AI-related fraud incidents, with 35% of UK businesses having fallen victim to such attacks, compared to 23% the previous year. This rise is attributed to the growing sophistication of tactics employed, including social engineering, identity theft, deepfakes, and voice cloning. These threats are beyond the capabilities of conventional multi-factor authentication and older voice biometric solutions.
Echoing concerns from AI experts, such as OpenAI CEO Sam Altman, who noted that AI has effectively “defeated” traditional voice biometric authentication, FARx introduces its next-generation biometrics technology.
Fusing speaker, speech, and face recognition, the software is adept at detecting synthetic and cloned voices. Its training, based on 55,000 synthetic voice samples from real telephony settings, empowers it to distinguish authentic human voices from AI-generated ones.
FARx 2.0 advances beyond standard voice user interfaces by identifying not only what is spoken but also who is speaking. This capability is crucial for thwarting identity spoofing via synthetic voices, deepfakes, or cloned audio and video. This innovation proves advantageous during customer onboarding, ensuring identities are neither fake nor synthetic.
Moreover, the system integrates smoothly into browsers, apps, and existing communication frameworks, offering a seamless multi-factor authentication process. Running unobtrusively in the background, it adapts to each user, discerning subtle biometric variations such as emotion and tone. This constant identity verification happens without disrupting user interaction and even captures biometric data from suspected fraudulent activities.
FARx 2.0 is particularly effective within Interactive Voice Response (IVR) telephony systems, enabling real-time detection of synthetic and cloned voices during calls in contact centres, helpdesks, and service desks. Additionally, it extends this security to video conferencing platforms for immediate deepfake detection.
The development of this sophisticated software follows FARx’s recent securing of £250,000 in seed investment, facilitated by the Seed Enterprise Investment Scheme (SEIS), which extends tax relief for investors who finance burgeoning startups.
Clive Summerfield, CEO of FARx, highlighted the continuous enhancement of their biometric multi-factor authentication technology: “This latest iteration of FARx is something we have been working on for a while now, with the aim to deliver an even more sophisticated, flexible biometric multi-factor authentication technology to users across a broad range of industries and applications. Receiving the investment through the SEIS has allowed us to do this at an even greater pace, speeding up the development and delivery of FARx 2.0 to those who need it most.”
In discussing recent developments, Summerfield noted, “In recent months, we have seen in real time, perhaps more than ever before, the true impact of social engineering. Data is already showing an increase in the use of AI for fraud and ID theft; as technology and AI develop, this kind of attack will only become more regular. Legacy voice biometrics and traditional MFA systems are simply no longer enough to outsmart the new era of AI-powered threats.”
