WebMink
Simon Phipps's personal commentary

Orwellian E-Tags

"Larry Downes teaches technology law and strategy at the University of California-Berkeley School of Information Management and Systems. He has no affiliation with the bar code industry." Larry is also either naive or engaged in an attempt at obscuring the real problem with RFID and its use with EPC in commerce when he writes in USA Today that we shouldn't fear "the new bar codes". He says:

The first generation of bar codes has helped do that for nearly 30 years. But if misguided privacy alarmists have their way, the benefits of the next generation of bar codes may be denied or delayed.

Misguided? Trying to call RFID tags carrying EPC codes "the next generation of bar codes" is a misuse of language of Orwellian proportions, using a known-harmless term to label a known-problematic technology. Over on the Interesting People list, Andreas Krisch makes this clear in his posting, which I'd recommend you read if you're not aware of the technology behind RFID, EPC and PML (there's a longer analysis in The Register). Andreas points out the real issue:

If the RFID-Tag is not destroyed or ­ better ­ removed at the checkout the consumer can easily be recognised by the EPC of her T-Shirt. With this unique identifier the retail shops are easily able to i.e. track the buying habits of their customers.

For 'T-shirt' read "any purchase" - the European Central Bank is even thinking of putting RFID tags in banknotes. The innocuous term 'bar code' seems wrongly applied to RFID/EPC/PML. To read a bar code I have to get access to the item. To read an EPC from an RFID tag I need just to get within radio range of the chip. Maybe a term like 'e-tag' for this technology grouping is more appropriate.

To be clear, the privacy problem with e-tags is the same as the one with national identity cards and other pervasive, public, unique IDs - triangulation. The problem of triangulation lays not in the nature of nor the intent behind the ID tags being used - e-tags, date of birth, social security number, vehicle registration - but in the ability to gather and cross-relate them with information gathered by other means. Any ID then becomes a 'key' on which to recall any of the other data.

Sadly, even the breathless activists Larry is criticising (not linked from the article as usual - CASPIAN and Stop RFID) don't really understand this it seems. The scheme itself is, as Larry points out, pretty innocuous in isolation, just like bar codes. Larry points this out when he says:

Many think of companies as amoral, profit-hungry beasts that will do anything to promote their own selfish interests. In the case of EPC, the early signs suggest an impressive cooperation aimed at making the transition as smooth as possible and of sharing the benefits of new technology as widely as possible.

But it's not the use that companies themselves will put the technology that's the problem. Unique EPCs that can be remotely read via radio and then checked against a server to retrieve XML-formatted summaries of their usage history provide the ideal means for orthogonal uses of the identifier. E-tags are not the only source of concern for those of us who believe privacy matters (and that includes my boss by the way - his quote asking you to "get over" lost privacy was in a context trying to make people face this very problem). But the ability to covertly gather EPCs from RFID chips that haven't been disabled or removed makes them much more worrying.

Here are some examples, all thankfully from my imagination so far.

• A political campaigner could read the e-tags of items carried by opposition supporters at a rally and then covertly track and undermine their activity.
• A paedophile could gather e-tags from clothes, sports gear and stuff near a school and then watch for passers-by elsewhere.
• Investigators could engage in covert 'electronic tagging' by accumulating the e-tags of items suspects owned and then use covert readers to track the suspect.
• TIA has not gone away, it's just hiding. E-tags could be a very handy hook on which to hang the work of this and MATRIX without needing a warrant (not that that seems to be much of a problem these days anyway)

So Larry does get one thing right - that consumers "must be able to permanently stop the transmission of data to or from tags once they leave the store." But this needs to be more than a voluntary guideline - it needs to be a requirement from the start. Larry finishes by saying "EPC isn't dangerous. Ignorance is." Larry, naivety is even worse.

posted at 10:35 PM (UK) | |