Benefit claimants to face lie-detector tests

Liberties groups warn of abuse of voice-analysis technology

The Government is preparing to test hi-tech voice analysis developed by the security industry in Israel and the United States, to catch out claimants who are not telling the truth.

The lie-detector technology, which measures stress, hesitation and other indicators of anxiety in a person's voice, will help identify people falsely pretending to be single parents or fraudulently claiming unemployment benefits.

Ministers believe the technology could be a valuable tool against fraudulent claimants, which cost the taxpayer £3bn a year.

Civil servants have been talking to insurance companies, which already use the technology to help uncover fraud in the UK, about the most effective way to test the voice analysis. They are looking at whether to use the technology in telephone conversations or one-to-one interviews.

The prospect of lie detectors being used by the Government has alarmed civil liberties campaigners who say it should not be used without the claimant's knowledge and needs regulation.

"New technology such as this needs to be carefully regulated to ensure the innocent are not swept up with the guilty. Surely there are more transparent methods for catching fraudsters?" said Jen Corlew of Liberty.

James Plaskitt, the work and pensions minister, admitted in questions from the Tories that he was currently considering how to pilot the lie-detector technology .

David Ruffley, the shadow minister for welfare reform, said he thought the technology could be a "valuable tool" against fraudsters. But he warned that it must be used with caution to stop it deterring genuine claimants from applying for money they are entitled to.

"There is still a huge problem with fraud and this software could prove a valuable tool. The total scale of fraud is totally unknown," he said. "That is why the National Audit Office refuses to sign off the Department of Work and Pensions' accounts. Some insurance companies use it, but the majority of UK insurers do not use it because they believe it will upset policy-holders.

"It is definitely worth looking at, but the question is are they looking at a large proportion of the claimants, a random range, or are they going to use it in a selection of case where there is a reasonable suspicion of fraud?"

Some insurers do not believe the technology is reliable enough. Unlike the polygraph, which measures blood pressure, heart rate and respiration, it measures stress in a person's voice in comparison to the answers to mundane questions such as "what is your name?".

Research has found that frequencies in the human voice can be affected by whether a person is lying or not. When a person is being deceptive the voice tends to get higher.

The technology is said to be able to differentiate between people who are anxious or hesitant and those who are lying.

A spokesman for the Department of Work and Pensions said: "We want to see how it would best work for us. We are looking at developing a whole process. It would be used across the board. It is definitely under consideration."