Complete and Utter Zebu

In an interview this weekend, Chancellor Alistair Darling referred to ID cards, saying there was probably no need for them because biometric passports, which carry the same information, would be sufficient. This has been taken by many as a sign that the project to get every Briton to have an ID card is likely to be abandoned, although it is by no means certain.

As we said only recently, it is appalling that authorities around the world are ignoring the fatal flaw in these systems, namely the problems any individual will face if their biometric identity is stolen, something that is remarkably easy to do. (See: ID cards have arrived – be afraid be very afraid)

There is, however, a glaring problem with biometric passports. Despite the impression given that they are perfection personified, the truth is that they are appallingly inaccurate.

The missing report
In 2004, the UK Passport Service conducted a biometrics enrolment trial. It published its report in 2005 but for some reason it can no longer be found on any government website. Given that the government is so keen on putting every one of its burps and farts online, we can only wonder why. Could it be that they don’t want us to read it and learn the truth? Surely not. Luckily, some people grabbed a copy while they had a chance (click here to read it).

The report found that, when it came to fingerprinting (using scanners, not the old-fashioned ink method), “the majority of participants achieved successful verification on fingerprint”. Sounds impressive. Until you look at the report and see that the actual figure was 81%. In other words, the system did not recognise 19% of people’s fingerprints, virtually a fifth.

Facial verification was worse, with a lamentable success rate of 69%, which dropped to 48% for disabled participants. In fact, one testing centre recorded a pass rate of just 6.25%, while one of the funniest parts of the report explains that operators at some centres held pieces of paper over lights to try to make it work.

In 2006, the Commons Science and Technology Committee urged further testing and consultation with IT professionals. Was its advice heeded? Like hell it was.

Live trial
In August 2008, Manchester Airport became the world’s first international airport to put biometric scanners into operation. They used facial recognition, because it is twice as fast as fingerprint verification. Not only was the technology not properly tested first but, amazingly, it was a “live trial”. Working on the deranged assumption that the system was secure, the scanners were not even manned by immigration officers. Anyone with a biometric passport could avoid the queues and pass through an unmanned gate.

Problems arose immediately. The scanners queried so many people that huge queues formed. So the machine were recalibrated or, in layman’s terms, fiddled. Rob Jenkins of Glasgow University, a world expert in facial recognition, said this made the system useless. He claimed that it could not distinguish between Gordon Brown and Mel Gibson or even between Winona Ryder and Osama bin Laden. In a subsequent interview for Hard Focus, Jenkins said:

“The real danger is in rushing to large-scale security deployments. For applications such as passport control or forensic face recognition…we know that the available technology is not yet up to the task…but the technology is simply being phased in, whether it works or not.”

Despite this, this far-from-foolproof system is about to be extended to include Heathrow, the busiest airport in the world. Nice to know that while troops are fighting and dying in Afghanistan, any terrorist with a biometric passport will find it a doddle to waltz into Britain through passport control.