I don't blame companies for hesitating to invest--biometrics systems still have problems, despite IBM’s prediction of advances. A prime example is how some fingerprint readers fell victim to the highly advanced gummy bear attack, in which a user acquires a gummy bear, applies it to the reader, and presses down. The sensor reads the fingerprint from the last user, which has now transferred to the gummy bear. The reader is defeated, the gummy-wielding attacker is authenticated as the previous user, and the system has become worthless. Organizations have been forced to replace hardware and software in light of this attack and revert to legacy methods, such as passwords, that are not vulnerable to rubbery candy.
More secure, it's hoped, are the digital images the government is embedding in the newest version of the U.S. passport for use with facial-recognition software, to reduce the likelihood of someone successfully using a fake passport to enter the country illegally. Since 2004, the US-VISIT--for United States Visitor and Immigrant Status Indicator Technology--program has been collecting digital fingerprint and facial images of international visitors to be used for identification; this data is shared with a number of government agencies. The enrollment and validation of these attributes is fast and accurate enough for use in everyday, large-scale deployments, and the Department of Homeland Security just announced it will pay Accenture Federal Services $71 million over 13 months to further improve the system.
Though they should, most users never question the privacy, storage, handling, and sharing of their biometric data. What happens if people are enrolled in a system and their biometric data is compromised, sold, shared, or mined in some way? This topic came to the fore in 2009 when a company offering faster airport security checks closed its doors and didn't immediately state where the biometric data it had collected would end up. In return for allowing Clear (which has since been reopened) to keep biometric data on file, frequent fliers could move through airport security faster. It was great for those who fly often and don't want to waste time. It would also be great for those who want to steal this data to impersonate a frequent flier, for either malicious airport activity or use elsewhere. If a credit card is stolen, it's easy enough to close the account and get a new card. Not so much for a new fingerprint.
While some people will always like to think they're targets of a vast international conspiracy looking to frame them for a failed government takeover, in reality, I don't see biometric data being targeted in such a way. On the other hand, this data could be sold to and mined by companies with the ability to analyze our physical traits, compare that to other data sets, store in-depth information about us, and perhaps disclose it all in some way that would harm us.
The fact that these concerns are mainstream shows that biometrics has evolved to a point where enrollment, usage, cost, and user fears are no longer hindering adoption. I can see a future in which governments push for inclusion of digital photos to be used with facial recognition, require fingerprints for traveling, and eventually embed DNA attributes in identification documents to address everything from fraud to immigration control.
As a user, it seems great not to worry about someone impersonating me and not having to carry an access token or know a password. At the same time, though, it's scary to think my fingerprint, DNA attributes, and digital image will be shared across governments, vendors, and employers. Those futuristic movies never addressed the security and privacy aspects of our personal biometric data and what happens if it's compromised, altered, or goes missing. That's up to us.
Adam Ely is security director at TiVo and a Dark Reading and InformationWeek contributor.