Unisys, which identifies itself as "a worldwide leader in integrating biometric technologies," on Tuesday released a survey that shows Americans happen to be amenable to the very biometric technology that it happens to offer.
Older and wealthier U.S.-based respondents indicated a preference for fingerprint scans as a method of verification. Seventy-six percent of those between the ages of 35 and 64, and 79% of those earning $50,000 or more annually, approved of this method.
Only 43% of respondents said they approved of hand and blood-vessel scans. They approved of other forms of authentication as follows: photographs (69%), personal identification numbers (69%), eye scans (61%), voice recognition (55%), and face scans (52%).
In a statement, Mark Cohn, VP of enterprise security at Unisys, characterized biometric technology as "an effective way to protect data and identities." He expects biometric technologies will be deployed more widely in the future at airport security checkpoints and banks.
According to Unisys spokesperson Danielle D'Angelo, the Unisys study included about 12,000 people worldwide, and just over 1,000 in the United States. She said that in countries like Australia and Malaysia, where biometric systems are more widely deployed than in the United States, acceptance is even greater.
Respondents were asked a single question about their willingness to trust various biometric technologies as an identity qualifier. "People are not that opposed to biometrics," she said. "Fingerprints and passwords are neck and neck. You hear that people are afraid of it, but I think there's a big myth there."
Perhaps people should be afraid, not because of the technology's Orwellian associations but because of its fallibility. For example, on Monday, Bach Khoa Internetwork Security (Bkis) published a report indicating that the face-recognition software that provides authentication on Asus, Lenovo, and Toshiba laptops running Windows Vista "can all be bypassed, even when set at highest security level."
To do so, an attacker would have to capture a picture of the user. Initiating a Skype video call, for example, could provide an opportunity do this. With some additional processing, the captured image becomes a key that the attacker can use to open the face-recognition lock that protects the user's laptop.
Biometric spoofing isn't a new concern.
In 2005, researchers at Clarkson University conducted tests to determine how easily fingerprint scanners could be fooled by Play-Doh molds of live fingers and fingers cut from cadavers. They tested more than 60 fake samples; the false verification rate was 90%. In previous years, other researchers had similar success with gelatin fingers.
In a recent note about a special issue of The International Journal of Biometrics devoted to biometric spoofing, editor in chief Khalid Saeed of Poland's Bialystok Technical University observes that spoofing affects even the most sophisticated systems. "The problem is that the faster the anti-spoofing technology grows, the more unexpectedly efficient spoofing techniques are," he explains. "When fingerprints were first spoofed, researchers started providing methods to detect liveness of the finger before fingerprint identification for personal verification. This led to the situation that spoofers started producing and using advanced special materials to simulate true-to-life texture."
Biometric systems, in other words, face the same arms race that other computer security technologies face.