Biometric authentication is seen by many to be the natural successor to passwords, but nobody has been able to figure out which method is the best. Fingerprint scanning is the most-popular option at the moment, with retina scanning coming in at a distant second, but some researchers from Binghamton University are working on a method that’s about as futuristic as they come: brain scans. They’ve developed a system known as Brainprint, which uses an electroencephalogram cap to scan their brains while they view 500 images in succession. According to the researchers, the system has a 100% accuracy rate so far, but obviously they’re going to need to refine the process significantly before it sees any serious adoption.
Passwords are on the way out: they’re difficult to remember, easy to crack, and offer up a relatively simple way for other people to get at your sensitive data. Imagine a prisoner being let out of the jail gates just because he knows the security phrase of the day – and not because he looks anything like a prison guard. That’s why technology firms and scientists are working hard on alternative solutions that are based around biometrics. The fingerprint sensor you might have on your phone is perhaps the best example: it takes no effort to ‘remember’ your fingerprint, but it’s very difficult to replicate or steal if you’re not who you say you are (Mission: Impossible films aside). Now scientists working at Binghamton University think they’ve found a suitable replacement for the password: a brain scan. The system they’ve developed, called Brainprint, uses an electroencephalogram (EEG) cap and then presents a succession of 500 images to the user at the rate of two per second – by studying the brain’s responses to these pictures, the associated software is able to pick out one person in 30 with 100 percent accuracy. No one’s about to start shipping laptops with EEG caps any time soon, but if this technology can be refined and improved upon, it could eventually provide an ultra-secure way of confirming someone’s identity that would be almost impossible to spoof (Tom Cruise et al., take note). Of course, for the system to work, it needs to know how you react to particular images in the first place so it can correctly identify matches.
