Britain is rolling out pilot programs to issue biometric ID cards in Newcastle, Glasgow and Leicester, and next month a mobile unit will begin visiting many British cities. It is estimated that 80% of British workers will have biometric ID cards by 2012, and the Home Office hopes to make them compulsory by 2014. However, 4% of iris scans -- 1 in 25 -- have been unsuccessful so far, and the UK government is apparently expecting a failure rate up to 7% (1 in 14). Concerns have been raised that the cards could worsen racism, on the apparent grounds that immigrants will be required to have the card before Britons will.
... in Germany the principle of "information self-determination" saw the federal constitutional court rule that although ID cards are in use, the public would not be given unique identity numbers that could be accessed by all government departments.
However even in Germany research has found that the police are more likely to stop people from ethnic minority groups to check their ID cards than the rest of the population [, thus] creating more potential for racial harassment and bad feeling between the police and ethnic groups.
The British government seems undecided on whether to market it as an anti-terrorist program or as control of illegal immigration, but pulls out the old, old argument, "What has anybody to worry about having their true identity known?" (Also known as, "Why are you objecting, if you have nothing to hide?")
Britain previously instituted a national identity card scheme at the beginning of WW2 in 1939; the government's continuation of the program after WW2 collapsed in 1950 when one North London man rebelled. The High Court ruled on the case in 1951 that continuation of the scheme was an annoyance to much of the public and "tended to turn law-abiding subjects into law breakers." Richard Davies, from the global watchdog Privacy International, said public opposition for an ID card in Australia shifted from 10% to 90% during a campaign in 1987, and New Zealand also backed off during the same period after strong public protests. Many European countries require ID cards, though.
"It didn't take much to persuade the entire Australian population," he said.
"All you need to do is present people evidence of the way authorities can abuse power," he said, adding that he believed the UK card would breach the European Convention on Human Rights.
The Earl of Selbourne, chair of the Royal Society's science in society committee, warns of hidden dangers with such biometric ID cards and centralkized records databases, saying that the British public is "sleepwalking into its future."
Meanwhile, Bruce Schneier's Crypto-Gram publishes a result from a Japanese mathematician who's been able to fool 11 different manufacturers' biometric fingerprint scanners 80% of the time using a few dollars' worth of readily available kitchen supplies and, with a little more effort, create a gelatin "fake finger" from a fingerprint that will also fool the scanners 80% of the time.
Update:
Charles Shannon Hendrix makes the following observations:
I worked in this industry briefly, and we joked all the time about how the customer was buying a false sense of security. For every product we worked on, we had a dozen ways to get around it, or a dozen ways you could abuse it. Our general consensus on biometrics and smart-cards was that it gained little for an individual. It's convenient, but does little and perhaps nothing to really protect you.
On the other hand, it can be a big benefit to large organizations, and is an amazingly easy thing to abuse. It's also perhaps the biggest and most dangerous Pandora's Box we've ever opened.
Biometrics are forever. You can't change your biometric identification like you can a numeric ID. Once the wrong people have it, they have it forever.
All biometrics should have a few requirements:
- All algorithms used to generate a key should be non-reversible so that keys generated for verification cannot be reversed to reveal the biometric data. This keeps you from being tracked, but you still get the benefits of biometrics to the individual. Large organizations and governments hate this idea of course.
- Biometric information should never be used unsalted: that is, it should never be used alone. This way nothing associated with your ID is permanent. Everything based on your ID is still very personal to you, but you can effectively change your biometric ID if you need to. Need is something that the individual should be free to determine.
- We should try to use biometric data that requires user consent. For example, fingerprints can be lifted easily, but retina scans generally require voluntary action. Fingerprints are a really bad way of doing it.
- Finally, there is rarely a NEED for things to have any security at all. Anonymous cash is perfectly feasible, and there is absolutely no reason this needs to be tracked. Yes, stealing it is like physical money in that you lose that money, but that's been true for over 2000 years now. It's at least not a new problem.
- When you do want security, them make sure as above that all keys are non-reversible, and transaction keys should be short-lived. Even if "they" record them, they will be invalid after the transaction is completed. The ONLY thing necessary is validation of the transfer of information, whether that be money, your order, whatever. There is absolutely ZERO need for transactions to be recorded.
Bottom line: we can use biometrics and also create perfectly secure transactions, and so so without privacy violations and abuse. The way things are going though, we are going to have to be far more vocal. Current proposed systems are full of holes, are insecure, and are a disaster for individual security and privacy.