Uncle Sams Got an RFID Jones

Opinion: I don't know what to be less impressed with: the arguments for RDID in security documents or the claims that it will always be a privacy disaster.

Governments seem determined to adopt RFID in identity documents and to view it as a security device. Privacy advocates seem determined to oppose them in all cases. I think this stuff is complicated.

Now the state of Washington has come up with a plan for a voluntary pilot project of drivers licenses that integrate many security features and radio-frequency identification. One of the big goals is to fit the requirements of the federal governments Western Hemisphere Travel Initiative. You may know this as the rules recently announced that would require a passport to travel to Mexico or Canada. In fact, the rules were more complex and didnt strictly require a passport.

The WHIT opened up the possibility of other ID that could meet the border-checking needs of the Department of Homeland Security. The "enhanced" Washington drivers license will cost more than a standard one and will require proof of citizenship, residence and identity. It will contain many (unspecified) enhanced security features found on passports.

/zimages/4/28571.gifLondons Royal Academy of Engineering suggests that someday a terrorist will be able to read personal details from a distance and set a bomb to go off when a particular person gets within range. Click here to read more.

Why does the government want to standardize ID? As the DHS said in its press release about the Washington drivers license pilot program, "At present, U.S. Customs and Border Protection personnel consider more than 8,000 distinct state issued birth certificates, drivers licenses or other forms of identification when making decisions on who and what to admit into the country."

This really is a legitimate concern. I know many people pooh-pooh the security value of more rigid standards for documents, but it makes sense to me. The more document types people can produce, the less scrutiny they can get.

And then theres the RDID component, the really controversial part. Im suspicious of the value of having RFID. If you take the security of the card seriously you have to have a person scrutinize the card and the person bearing it. RFID lets the authorities easily bring up records of who should own that card, but so could a 2-D bar code. My New Jersey drivers license has one of these bar codes on it.

Back when I wrote about the new ePassports the question seemed simpler, even if the opposition was just as hysterical. But there was a crucial difference: The ePassport has a chip that transmits all of the ID information in the passport, including the photograph. This makes it easy to conceive of privacy and other security breaches.

I quoted Kevin Ashton, the co-founder of MITs Auto ID-Labs, which gave birth to EPCglobal, the international network for tracking items through a supply chain using RFID. EPCGlobal defines a key numbering system, which is implemented in practice by VeriSign, so that items can be tracked throughout the supply chain worldwide. Readers can read the chips anywhere and use the number as a database key for lookups, or simply report it on to some database for tracking of its movements.

Ashton argued that if RFID is in passports at all, it should be implemented the way EPCglobal does it—all the chip stores is a unique code. You need to have access to the database to know anything about the holder of that code. This is what Washington has done: The only thing that the drivers license RFID transmits is a code.

Its possible to imagine abuses of even this form of RFID simply through tracking. If you string readers all over the place and they can actually read the cards, which normally should have a very short range, then you can track the movements of people. The DHS fact sheet on this technology points out, among other things, that "all card holders would be issued a protective sleeve for the card, preventing transmission of the Vicinity RFID signal while the card is in the sleeve," but Im not sure people (especially men) would keep such a sleeve on their license in their wallet. (Women have many more options for RFID privacy.)

Personally I think this complaint is far-fetched. Theyd need to get you to produce the card to have any confidence in such a program of tracking, and at that point its no less safe than a bar code that they could swipe.

In fairness to the passport people, they argued (not persuasively enough I would say) that passports must be read all around the world, in places where access to the database couldnt be assumed. Too bad I would say; things will have to be slow in such places, pretty much like theyve always been. And its not like you want, for example, the Venezuelan Foreign Ministry to have access to the database.

But in the state of Washington it seems to me the trade-offs are good ones. If we accept it as a proper government function to issue ID and to check it, then its not unreasonable to allow them to do a good job. Government abuse of the potential of RFID can be dealt with in legislation, and its part of governments job to block abuses by third parties. Weak ID is not a good answer.

Security Center Editor Larry Seltzer has worked in and written about the computer industry since 1983.

More from Larry Seltzer

/zimages/4/28571.gifCheck out eWEEK.coms for the latest security news, reviews and analysis. And for insights on security coverage around the Web, take a look at eWEEK.com Security Center Editor Larry Seltzers Weblog.