The insurance industry hemorrhages over $308 billion annually to fraud, a staggering figure that translates to roughly $900 in additional premiums for every American family. Traditional verification methods, from document reviews to lengthy interviews, struggle to keep pace with increasingly sophisticated schemes. But what if the solution wasn't more paperwork, but rather the human voice itself?
Scott Moore, an executive at Clearspeed, a BIG Innovation Award winner whose technology is reshaping how organizations separate truth from risk. What Moore revealed challenges everything we think we know about detecting fraud.
The Science Behind Voice Screening:
Clearspeed's approach sounds almost science fiction: measuring neurophysiological responses through voice characteristics to detect deception. But Moore was quick to ground the conversation in biology.
The technology captures what happens when humans face fight-or-flight moments—those instinctive responses hardwired into our nervous systems. When someone answers a yes-or-no question about their claim or application, specific vocal characteristics change if they are withholding information or not being upfront.. Clearspeed's system isolates these changes, requiring just 300 milliseconds of voice to generate a risk assessment.
Moore emphasized this isn't traditional lie detection, and it's not trying to read emotions or analyze speech patterns the way AI typically approaches voice. The system measures the presence or absence of a specific biological reaction—one that humans have no conscious control over–and that correlates with risk. That's a crucial distinction.
Accuracy Changes the Game.
The numbers Moore shared were striking. Clearspeed validated its technology against Department of Defense ground truth data, achieving 97% accuracy in detecting the neurophysiological response. But perhaps more impressive is what he called the 100% measure—the system always detects whether that biological reaction is present or absent. The only variable is what triggered it.
This matters enormously for fraud detection in practice. Clearspeed functions as a triage tool, generating red, yellow, and green risk bandings. Green respondents—those showing no stress response—can be fast-tracked through normal processes–that’s usually about 90% of people in a process. The system then concentrates investigative resources on the smaller pool of elevated-risk cases.
Moore offered a memorable analogy: Clearspeed works like an MRI machine. It detects anomalies, but a doctor still makes the diagnosis. The technology clears the hay, allowing organizations to work the needles—a significant shift from traditional approaches that treat every claim or application with equal suspicion.
Eliminating Bias from the Equation
One of the most compelling aspects of Clearspeed's approach is what it deliberately doesn't do. The system captures no personally identifiable information, learns no historical patterns, and creates no demographic profiles. Moore called it the "least radioactive data insight" an organization can add to their infrastructure.
Traditional fraud detection often relies on pattern matching—flagging claims that resemble past fraudulent ones or applicants who share characteristics with previous bad actors. This approach inevitably embeds historical biases into decision-making. Clearspeed sidesteps this entirely by measuring only the biological response, blind to everything else about the person.
Moore reported that testing across multiple demographics—age, gender, ethnicity, and education level—showed no variation in the system's efficacy. The technology works identically across all populations, addressing a persistent concern in AI-powered decision systems.
From Battlefields to Insurance Applications
Clearspeed's origins explain its unusual capabilities. The technology was developed to solve a critical military challenge: vetting local assets supporting U.S. operations overseas. Traditional polygraph examinations took months. Going faster meant accepting unacceptable risk. The solution needed to be language-agnostic, voice-based, and capable of maintaining accuracy at scale.
That military DNA now translates across what Moore described as "10,000 potential use cases." Insurance carriers deploy Clearspeed across the entire customer lifecycle—from application and underwriting through claims and fraud investigation. Banks can use it for transaction verification. Government agencies can apply it to Medicare and Medicaid fraud, tax filing verification, and social security benefits.
The Real Challenge: Changing How Organizations Think
Despite delivering what Moore called "50X ROI," Clearspeed faces an unexpected obstacle: organizations struggle to reimagine their processes around new capabilities.
Consider fraud detection in auto claims. If a carrier assumes 3% of claims are fraudulent, they've designed their workflow around that assumption. When Clearspeed reveals the actual number might be 10%, the organization must fundamentally redesign how they consume that intelligence. The insight is valuable, but using it effectively requires starting from a blank whiteboard.
Moore put it bluntly: "Selling Clearspeed is easy. Consumption and illumination is hard." Organizations must overcome decades of institutional assumptions about what's possible—and affordable—in verification. When a new tool proves that expensive external verification can be replaced with a quick voice assessment, the implications ripple through entire departments.
Looking Ahead: Trust at Scale
As synthetic voice threats and deepfakes proliferate, the need for genuine voice-based trust signals intensifies. HSBC has already enrolled over 15 million customers in voice biometric systems, reducing fraud attempts by 50%. But Clearspeed offers something different—not voice recognition, but voice technology that reveals risk and can build trust.
Moore's advice for other innovators navigating adoption challenges came down to one word: focus. Despite suggestions to apply the technology to dating apps, academic integrity, and countless other applications, Clearspeed maintains strategic discipline in its core markets. Financial transactions, fraud prevention, and national security represent enough opportunity without chasing every possible use case.
For an industry losing hundreds of billions annually to fraud while simultaneously raising premiums on honest customers, voice screening offers a compelling path forward. The technology exists. The validation is proven. The question now is whether organizations will embrace the blank whiteboard—and build verification processes that finally work at the speed of trust.
Catch the full conversation with Scott Moore on The Winners' Circle podcast, and discover how BIG Innovation Award winners like Clearspeed are transforming their industries.