VoxEQ, a provider of AI tools for voice biometrics, has announced a breakthrough in “age from voice” prediction, besting the state of the art for accuracy by twofold.

“The use of voice biometrics and biomarkers has become a mainstay with large banks, insurers, online platforms and others fighting fraud, account takeover (ATO) and identity theft,” said Jack Caven, CEO of VoxEQ. “In the sectors we operate in, take banking as an example, fraud and ATO losses are larger than ever and growing faster than before – some $1 billion in losses every week. Our clients and partners tell us that improving accuracy is not only critical to achieve the underlying goal of catching fraudsters before they cause harm, but that the cost of poor accuracy on customer experience, call center operations and brand is unsustainably high.”

According to the company, the top 15 US banks receive a projected 44 million voice calls per day. Some 29,000 of those calls are known fraud, identified after the fraud occurred. However, fraudsters have gotten more sophisticated in their tactics – to include use of the dark web for personal information and technology workarounds to bypass authentication tools. Thus, many of the tools and products available today are losing their efficacy.

VoxEQ provides a layer of preventative security via biometrics that, on its own, or conjunction with other tools, enhances real-time voice fraud identification upfront – and doing so in a privacy protecting manner.

“For VoxEQ, surpassing state of the art accuracy in age, gender, voiceprint and other key measures has been the number one goal since our launch,” added Caven. “Achieving that goal and deploying it in our flagship product, ATOM (Account Take-Over Mitigation), means that our cloud-based solution has the potential to stop far more fraudsters with less friction for customers and fewer risks and mitigation costs for institutions.”

VoxEQ layers two unique, proprietary biometric features on top of its core capabilities; “Dynamic Confidence”, a determination based on unique factors of each voice occurrence, such as the quality of the signal, that informs a broader or narrower range of conviction for the insight provided and “Customized Acuity”, a feature that allows clients to adjust the tool’s sensitivity to minimize fraud costs within their own unique objectives, constraints and context.

“More fraud and security professionals are looking for flexibility and accessibility in the solutions they deploy. They want them built for scale, using an API approach, fully integrated into the systems and platforms they already leverage – and with a high level of customization,” said Caven. “VoxEQ gives companies these operational degrees of freedom, superior detection and responsiveness to today’s threats and better footing to address those coming over the horizon.”