Esure stressed that it "will never treat VRA analysis as 'evidence' of potential fraud but will use it as a tool to direct the examination of a particular claim towards hard evidence that can be found through focused investigation."
"This," says esure, "will help to ensure that `false positives' - where a genuine claim might be turned down - never occur using the system."
The company also emphasised that its technology differs from lie detection. "VRA cannot tell the content of someone's speech only the frequency of that speech," it explains.
It has been shown that when a speaker experiences stress when answering a question or recounting an exaggerated or false statement, that the frequency of his voice changes. It is this that VRA registers and assesses.
The system is intended to 'fast-track' cases where no adverse risk is registered during the claim logging call, allowing payments to be made more quickly. But where a high risk is indicated during the call, further investigation will be carried out.
The VRA system will be piloted by esure for six months – initially for higher risk claims, such as unrecovered thefts and burn-out car claims. In the future, this is likely to expand to potential staged accidents and home insurance claims.
Gordon Hannah, esure's Head of Claims, said:
"This system is not intrusive. Claimants will be told clearly that their call is being recorded and analysed. We believe the honest majority is fed up with funding the cost of fraud so it is essential that esure works to identify and act against potential fraud. Voice risk analysis is just one tool, but we believe it is a very strong one."
To alleviate concerns about privacy implications of the system, esure says it is writing to all UK civil liberties and consumer groups and asking them to contact the company if they have any concerns over VRA.