At 07:48 AM 7/25/2006, David Farber wrote:
In case anyone needed more proof that we're all living in a Philip K. Dick novel, a pair of hackers have recently demonstrated how human- implantable RFID chips from VeriChip can be easily cloned, effectively stealing the person's identity. ... For its part, VeriChip has only said they haven't yet had a chance to review the evidence but still insist that "it's very difficult to steal a VeriChip."
Certainly literally true, if by "steal" one means, "get one's hands on the original, e.g., pry one out of Annalee Newitz's arm." But we should recongize that the vast majority of RFID applications [BUT NOT ALL djf] don't depend on inability to clone them. RFID tags in most commerce will be as unclonable as license plates, which anyone with a little tin, paint and shop skills could zap out copies of, but which nonetheless serve as a cheap means for reasonably reliable identification. Think of most RFID applications as just like print bar codes; there have been various cases of fraud committed against systems employing the latter, most notably where thieves use bar codes for inferior goods to purchase expensive ones ("Bar code says that's a drill bit, and it looks like a drill bit...") then return the goods to pocket the difference in price. The new wrinkle that RFID offers for commerce here is uniqueness: the local Home Depot currently knows that it has 500 units of carbide drill bit, all bearing identical bar codes... in an item-level RFID tagged world, it would know 500 unique serials, so spoofing the checkout clerk with a false tag becomes a little harder. And, with unique tags, it becomes easier to compile and retain longitudinal dossiers on "where has this thing been?" (if the various parties in supply chain actually read the tags): this is the aspect that will be used for pharmaceutical knockoff detection, where the overarching RFID tracking and management system will be able to provide some provenance information ("This very bottle was allegedly seen in Singapore 3 hours ago... something's not right"). This is also one of the more privacy-invasive aspects. I've seen one research effort (an NSF SBIR) looking at creating unclonable RFID thus far, which basically works, I believe, by extracting a physical signature of the item to be tagged (in the awarded research, it was magnetic signatures), and using that as part of the unique key, or perhaps registering that signature in an off- chip database that would need to be additionally queried. In the VeriChip hack, you might address the problem that that little chip merely spits out a unique ID that anyone who can read can rewrite into a new chip by having the implanted chip also encode some (relatively) unclonable aspect of the person the chip is embedded in, e.g., you can still "steal" the unique ID, but could only then use it in a chip in another (1) female; with (2) brown eyes; (3) blood type AB-; etc.; etc. But so far as I know the VeriChip used in human implants is just that little unique number... its value as a unique ID for security authentication depends a lot on it being hidden from 3rd party readers. Of course, we have this problem in spades all over the place... your SSN, or credit card number, can be fairly easily abused by anyone who knows it, despite the fact that you have to expose it to a lot of parties, many, many times over the course of a year. Ross ---- Ross Stapleton-Gray, Ph.D. Stapleton-Gray & Associates, Inc. http://www.stapleton-gray.com http://www.sortingdoor.com ------------------------------------- You are subscribed as eugen@leitl.org To manage your subscription, go to http://v2.listbox.com/member/?listname=ip Archives at: http://www.interesting-people.org/archives/interesting-people/ ----- End forwarded message ----- -- Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE [demime 1.01d removed an attachment of type application/pgp-signature which had a name of signature.asc]