Europe has adopted a new radio-frequency identification (RFID) privacy framework.
It won't come into effect for six months, and even then will be voluntary, but it does attempt to put the European citizens' privacy onto the RFID agenda.
The "Privacy and Data Protection Impact Assessment Framework for RFID Applications" was published in January, having been in production for six months, and is now endorsed by the European Network and Information Security Agency (ENISA) as the standard to which RFID deployments in Europe will be held to account – if they wish to be.
The document (24-page PDF/257 KB, and remarkably soporific) was put together by the RFID industry – which recognises that public perception is against it – and the Article 29 Working Group, a committee of national data protection chiefs. Despite being voluntary it will be widely adopted, partly because compliance isn't particularly onerous and partly because public fears need allaying.
The Framework identifies four levels of RFID application, requiring different quantities of scrutiny. Level 3 is where personal data is stored on an RFID tag itself; Level 2 has the tag holding a database key linked to personal data. Level 2 and 3 both require a full audit.
Level 1 systems aren't linked to a person, but might be carried by a person and so require a mini-audit. Finally, Level 0 systems are attached to pallets, crates and so forth and have no privacy implications.
The audits, or Privacy Impact Assessments (PIAs), to use the Framework's nomenclature, are carried out by the "RFID Application Operator". The PIA includes such things as checking each stage of transmission is suitably encrypted, thinking about how (and why) a miscreant might attack the system and how the system might be illegitimately utilised – basically the stuff any decent systems architect would do by habit.
The PIA is then presented to "the competent authorities"... perhaps; it is up to member states to decide who is competent and if presentation of a PIA is mandatory or simply has to be held by the deploying company for production on demand (realistically unlikely, until something goes wrong).
The Framework might help large companies focus on the privacy implications of their RFID deployments, and that's useful, but as every phone becomes an RFID Tag reader we can't help wondering if small-scale deployments won't be at greater risk of privacy abuse. Such systems will soon become incredibly cheap and easy to deploy, and will be open to exactly the kind of feature creep that creates security problems.
The EU Framework might prevent a major supermarket tracking the clothes you buy, but it won't stop the local gym recording which exercise machine you've been using, unless the local gym chooses to play along. ®